Oilswelling Cement
Oilswelling Cement
Oilswelling Cement
h i g h l i g h t s
An oil swellable polymer (OSP) with oil absorptivity as high as 700% is synthesized.
Addition of OSP decreases the total porosity of the hardened cement paste (HCP).
Addition of OSP reduces the water adsorption rate and increases the oil absorptivity of HCP.
Incorporation of OSP significantly increases the expandability of HCP upon oil absorption.
A novel technique to achieve a crack self-healing property is developed.
a r t i c l e i n f o a b s t r a c t
Article history: This paper proposes a novel approach to achieve self-healing property of cement sheath in the oil well
Received 2 June 2015 triggered by oil leakage through cracks. An oil swellable polymer (OSP) with oil absorptivity as high as
Received in revised form 1 February 2016 700% was synthesized through emulsion polymerization. The OSP dispersion was added into the
Accepted 25 March 2016
Portland cement paste. It is found that the oil absorption and the linear expansion of the OSP modified
Available online 31 March 2016
hardened cement paste (MHCP) are greatly increased compared to the blank hardened cement paste
(BHCP). Pore structure measurements by N2 adsorption show that the addition of OSP decreases the total
Keywords:
porosity of MHCP which confirms that the formation of the polymeric film in MHCP. The self-healing
Oil swellable polymer
Oil absorption
ability of MHCP was evaluated by the oil flow test through a single crack in the cylindrical specimen.
Portland cement paste The oil flow tests indicate that the leakage of oil through the crack is gradually mitigated due to the
Expansion expansion of the MHCP and thus the self-healing ability is successfully achieved.
Ó 2016 Elsevier Ltd. All rights reserved.
1. Introduction paste, namely the cement sheath, fulfills many functions, including
mechanically bonding and supporting the casing, chemically pro-
Cracks, caused by shrinkage and external loads, are unavoidable tecting the steel casing from corrosion, isolating fluids and gases
in the cementitious materials because of their intrinsic fragile between formations. Consequently, the sheath needs to withstand
property [1]. These cracks facilitate the ingress of aggressive and various stresses and erosions during the service life of an oil well.
harmful substances into the cementitious materials and indeed Cracking inside the cement sheath and debonding between the
reduce the functionality and durability of the materials [1,2]. In cement sheath and the borehole (or the metal pipe) are often tak-
oil well cementing, cement slurry is used to cementing a metal ing place. The consequences of the occurrence of such cracks and
pipe (so-called casing) into the borehole. The hardened cement gaps are: 1) leakage of oil/gas fluid and thus reduction of the pro-
ductivity; 2) failure to protect the steel pipe from corrosion due to
the invasion of aggressive species such as saline fluid, CO2 and H2S.
⇑ Corresponding author at: Department of Civil Engineering, Tsinghua University, Detection of the occurrence of such cracks and manual repair dur-
Haidian District, Beijing 100084, China.
ing operation of the oil well are almost impossible because these
E-mail address: kxm@mail.tsinghua.edu.cn (X. Kong).
http://dx.doi.org/10.1016/j.conbuildmat.2016.03.163
0950-0618/Ó 2016 Elsevier Ltd. All rights reserved.
Z. Lu et al. / Construction and Building Materials 114 (2016) 98–108 99
cracks can be located as deep as beneath the earth’s surface in swellable polymer nano-particles were incorporated into HCP by
depth up to 9000 m. In this case, the only conceivable approach mixing the polymer latex into the cement slurry. For the fully
for the leakage problems caused by cracking is self-healing of the cured HCP samples, their oil absorption and their linear expansion
cracks in the hardened cement paste (HCP) triggered by the leak- upon oil absorption were measured. The flow behavior of oil fluid
age of fluids, either aqueous fluids or oily fluids. through a single crack penetrating the cylindrical HCP specimens
Self-healing of cementitious materials, such as concrete in nat- was studied to signify the self-healing behavior of the polymer
ural environment was firstly reported by the French Academy of modified HCPs.
Science in 1836 [3]. The main mechanisms can be summarized as
followings [4]: 1) Further hydration of the unreacted cement or 2. Experimental
cementitious material and the hydration products are able to block
2.1. Materials
the cracks; 2) Formation of calcium carbonate on the surface of
crack through the carbonation reaction of hydration product Ca Analytical grade of chemicals, styrene (St), butyl acrylate (BA), octadecyl acry-
(OH)2; 3) Blocking cracks by impurities in the water fluid or by late (OA), divinylbenzene (DVB), sodium persulfate (SPS), sodium hydroxide acry-
loose concrete particles resulting from crack spalling; 4) Expansion lamide (AM) and methacrylic acid (MAA) were used as received (all >98% purity).
of the hydrated cementitious matrix. Such self-healing phenomena Anionic emulsifier MS-1 (alkylphenol ether sulfo succinate sodium salt, C8H17C6H4
O(CH2CH2O)10COC3H6-SO3Na, 40 wt.% aqueous solution) and non-ionic emulsifier
are called autogenous self-healing because no additional compo-
OP-10 (alkylphenol ethoxylates, C8H17C6H4O(CH2CH2O)10H 40 wt.% aqueous solu-
nents is needed in concrete formulations [3]. It is clear that water tion) were provided by Haian petrochemical factory. Dispersing agent (water-
or moisture is one of the necessary elements in all autogenous self- soluble sulfonated acetone-formaldehyde resin) and fluid loss agent (modified
healing reactions. However, in case of oil well, water is often polyvinylalcohol) used in the Portland cement paste preparation were provided
by the Sinopec Research Institute of Petroleum Engineering. Deionized water
absent. Therefore, the autogenous self-healing behaviors of oil well
(DI water) was used in all experiments in this study including the synthesis of poly-
cement sheath cannot be realized in the absence of water (or mois- mers and the preparation of cementitious mixtures. Diesel was chosen as the oily
ture) based on the abovementioned mechanisms. It is demanded to medium. The physical properties of diesel were shown in Table 1.
develop a novel concept to achieve self-healing properties of HCP According to Chinese standard GB10238-2005, G class oil well Portland cement
triggered by oil (or gas) leakage. produced by Sichuan Jiahua cement plant was used in this study. The chemical and
mineral composition of this cement was measured by X-ray fluorescence spectrom-
Besides the autogenous self-healing properties of cracks in
etry (XRF, LAB CENTER XRF-1800, SHIMADZU), and the mineral compositions were
hardened cementitious materials, the other concept for self- calculated according to the Bogue equation, the results were shown in Table 2.
healing is to introduce healing agents such as curable polymer
resins (epoxy resin) or sodium silicate in the forms of microcap- 2.2. Synthesis and characterization of the oil swellable polymer
sules, fibers etc into cementitious materials [5–9]. The healing
agents can be released and cure the cracks when the capsules or 2.2.1. Synthesis of the oil swellable polymer
It has been well documented that the oil absorptivity of polymer could be
fibers are broken. As a result, the cracks are healed. In this case, highly increased by incorporating long alkyl chains into the polymer, which can
the self-healing is triggered by mechanically breakage of the make the resulting polymer more hydrophobic [21]. In addition, molecular weight
microcapsules or fibers. Proper design and control of the microcap- (Mw) as well as the crosslinking density of the polymer plays an important role in
sules or fibers are keys to the success of self-healing. It is not an the oil absorption capacity of the polymer. High Mw and appropriate crosslinking
density are beneficial to high oil absorptivity [22]. Taking all these factors into
easy work to keep the microcapsules or fibers intact during mixing
account, we synthesized the oil swellable polymer by co-polymerizing styrene,
of the cementitious materials and to properly activate the curing butyl acrylate and octadecyl acrylate as a functional monomer via radical polymer-
process of the healing agents at the location of cracking. The intro- ization. The oil swellable polymer was prepared by emulsion polymerization in a
duction of epoxy resin without any hardener into cementitious 1000 mL three-neck glass flask equipped with a mechanical stirrer and dosing units
materials to achieve the self-healing is recently studied [10–13]. for both the monomers and the initiator. A water bath with fixed temperature of
90 °C was used to ensure the constant temperature during polymerization. Initiator
They concluded that the liquid epoxy resin remaining in the pores solution was prepared by dissolving 3.6 g SPS in 252.0 g DI water. Monomer emul-
of HCP can be released and hardened under the catalysis of Ca sion was prepared by mixing monomers (120.2 g St, 180.0 g BA, 148.5 g OA, DVB
(OH)2 and thus cure the microcracks. However, this method is 1.7 g), emulsifiers (3.6 g MS-1 and 3.6 g OP-10) and 540.0 g DI water using a
practically inapplicable in oil well condition due to the high solu- high-shear mixer at 10,000 rpm. Firstly, 225.0 g DI water and 12.8 g initiator solu-
tion were charged into the flask. The pre-charge was then heated up to 90 °C under
bility of epoxy resin in the oil. On the other hand, especially, in
stirring and kept for 5 min. The monomer emulsion prepared above and the rest ini-
the oil well cementing work, the cement slurries are prepared tiator solution (242.8 g) were then separately dosed into the flask at constant dos-
under super-high shear up to 10,000 rpm and the microcapsules ing rates. The dosing time of both the monomer emulsion and the initiator solution
or fibers are hardly able to withstand such high shear force during were 4 h. After that, the flask was kept at 90 °C for another 0.5 h to complete the
the mixing process. polymerization. Afterwards, the obtained polymer dispersion was cooled down to
the room temperature. Solid content of the prepared polymer dispersions was mea-
Another emerging self-healing concept for cementitious materi- sured by drying a certain amount of the polymer dispersion at 60 °C until constant
als in recent years is to involve a biological repair technique by the weight was reached. The measured solid content of the prepared polymer disper-
introduction of bacteria into concrete to aid in the precipitation of sion was 27.9%.
calcium carbonate to achieve crack self-healing [14,15]. However,
the oil well cement sheath is subjected to high temperature up 2.2.2. Characterization of the oil swellable polymer
Particle size of the polymer dispersion was determined by dynamic light scat-
to 200 °C and high pressure up to 70 MPa in deep oil well and
tering (DLS) with a Malvern Zetasizer 3000hs (UK). The chemical structures of the
almost no biological species could survive in such severe polymers were characterized by FTIR (Nicolet 6700 FTIR, Thermo Fisher Scientific,
conditions. USA) instrument with a resolution of 4 cm1. It is well understood that the latex
Apart from introducing the curable healing agents or bacteria to may contain a tiny amount of residual monomers, oligomers, and free surfactants
achieve self-healing, the expansion of hydrated cementitious
matrix will also benefit to the self-healing of cracks. Calcium sul-
Table 1
foaluminate based expansive agents [16–19], super absorbent Physical parameters of diesel.
polymers (SAP) [20] were studied to achieve the self-healing ability
of HCP through the dimensional expansion of HCP. Following these Viscosity at Surface tensiona at Density at Density at
25 °C 25 °C 25 °C 80 °C
ideas, this paper develops a novel concept to achieve self-healing of
HCPs stimulated by the leakage of oil fluid through cracks. Polymer 0.00435 Pa s 27.8 mN/m 0.846 kg/m3 0.827 kg/m3
Table 2
Chemical and mineral composition of G class oil well cement (wt.%).
[23,24]. The latex was purified by technique of dialysis in order to minimize the 2.3.3. Pore structure analysis
negative influences of those components on the FTIR spectrum of the major poly- Pore structure of the hardened cement pastes was analyzed by techniques of
mer of the latex. mercury intrusion porosimetry (MIP) and N2 adsorption. MIP has been one of the
To remove the components in the aqueous phase of the latex, a cellulose ester most commonly used methods to characterize the pore structure of the cementi-
semi-permeable membrane MD44-7 with nominal molecular weight cut-off of tious materials [27,28], although it has been often criticized majorly due to the mis-
7000 Da provided by Beijing Ruida Henghui Science and Technology Development interpretation of the received data [29]. In addition, in the polymer modified
Co. Ltd, was used in dialysis experiments, which allows globulin with a molecular cementitious materials, the polymer incorporated in the hardened materials may
weight of less than 7000 to pass through. A certain amount of latex was poured into form films and block some capillary pores for invasion of water or gas. However,
a dialysis bag and then immersed into 500 g DI water. The conductivity of the dia- the polymeric film may also be destroyed by the intrusion of mercury under high
lysate was continuously measured to estimate the cleanness of the dialysate. The pressure during MIP test [30]. Therefore, in this study, N2 adsorption method is
dialysate outside of the dialysis bag was renewed every 6 h. The dialysis ended employed to provide additional information of the hardened cement pastes in the
when the conductivity of the dialysate dropped below 30 lS/cm. The polymer film presence of polymer.
was prepared by drying a certain amount of the purified polymer latex at 60 °C for The HCP specimens were firstly placed into an alcohol bath (analytical grade) to
24 h and then scanned by a FTIR apparatus in the absorption mode from 400 to stop cement hydration [31]. After 3 d storage in an oven at 60 °C for completely
4000 cm1 at 25 °C. drying, the specimens were cut into small pieces of diameter about 10 mm and
The water and oil absorption capacity of the dried polymer were tested by thickness 5 mm for MIP test using Hg-porosimetry (Autopore, IV 9510, USA). Other
immersing the polymer film into oil and water bath until a constant weight was pieces of the specimen were ground into powder and additional vacuum dried at
reached. The water or oil absorptivity was calculated based on the following least 3 h for N2 adsorption test using a surface area and porosity analyzer (TriStar
equation: II 3020, Micromeritics, USA).
R ¼ M 1 =M 0 100% ð1Þ
2.3.4. Scanning electron microscope (SEM)
R stands for the water or oil absorptivity, M1 represents mass of the polymer film SEM was employed for morphological characterization of the hardened materi-
with absorbed water or oil, M0 is the initial mass of the dried polymer film. als. SEM imaging on the fractured section of the BHCP and MHCP were conducted
using the FEI Quanta 200F scanning electron microscope under high vacuum. Prior
to SEM imaging, the fractured surface was sputtered with carbon.
2.3. Paste experiments
2.3.2. Absorptivity and linear expansion measurements 2.3.7. Oil flow test trough a single crack in the HCP specimens
The obtained HCP specimens were completely dried in an oven at 80 °C for In order to check the self-healing of cracks in HCP specimens in case of oil leak-
about 7 d to reach a constant weight and a constant dimension due to the loss of age, oil flow test trough a single crack penetrating the HCP specimens was con-
the free evaporable water in HCP specimens. The dried specimens were then imme- ducted by using an apparatus as shown in Fig. 2. A cylindrical HCP specimen
diately immersed into a water (or oil) bath with constant temperature of 80 °C for with a diameter of 10 cm and a height of 3 cm was prepared by following the same
water (or oil) absorption tests. The specimens were weighed every 30 min at the procedure as described before. The HCP specimen was then split into two parts by
first day of immersion and later every 3 h to follow the mass change of the speci- using the device for the three point flexural test. The two parts of the specimen
men during immersion. The absorptivity measurement was continued for 14 d. were then assembled and tightly fastened using a steel frame with a rubber ring
The dimensional change of the HCP specimens immersed in water/oil bath was for sealing all gaps to avoid possible leakage of fluid through gaps. Thus, an artificial
immediately measured after drying at 80 °C by using the apparatus shown in Fig. 1. single crack penetrating the whole HCP specimen, which is geometrically similar to
The vertical expansion of the HCP specimen under water/oil immersion was contin- the naturally formed cracks, was generated for a fluid flow test. The preparation of
uously measured and the temperature of the water/oil bath was controlled con- the HCP specimen with a single penetrating crack is schematically shown in Fig. 3.
stantly at 80 °C in a temperature chamber. The measurement was ended when The oil flow test was conducted on the pre-cracked HCP to investigate the crack
the vertical dimension of the HCP specimens stay unchanged. self-healing property in case of oil leakage through the crack. In this study, diesel
For each HCP sample, both the absorptivity and linear expansion measurements was chosen as the fluid for the flow test. The oil flow through the crack was driven
were repeated three times and a representative result was used for discussion. by an additional pressure in the oil tank provided by high-pressure nitrogen vessel.
Table 3
Mixing proportions of the cement pastes.
Samples Oil well cement/g Water/g Fluid loss agent/g Dispersing agent/g Polymer dispersion/g
Blank HCP (BHCP) 1600 704 64 4.8 0
Polymer modified HCP (MHCP) 1600 291 64 4.8 573
Z. Lu et al. / Construction and Building Materials 114 (2016) 98–108 101
Fig. 1. Apparatus for the measurement of the vertical expansion of HCP specimens immersed in water/oil bath.
Fig. 2. Schematic drawing of the apparatus for the oil flow test.
The pressure difference between the inlet and the outlet of the crack was varied Section 2.2, FTIR analysis was performed on the synthesized poly-
from 0.02 MPa to 0.12 MPa. Under a certain pressure, the mass flow rate of the fluid
mer. Prior to FTIR analysis, the obtained polymer dispersion was
was recorded for 1 h to obtain a stable flow curve.
After the flow test under varied pressures, the HCP specimens were then stored
cleaned via dialysis technique to remove all residual monomers,
in the oil bath for another 2, 4 and 6 d respectively and the flow test was repeated at free surfactants and salts in aqueous phase and then completely
each storing age to check the change of the crack during oil immersion. dried to get rid of all residual monomers contained in the OSP poly-
mer. The FTIR spectrum of the cleaned OSP polymer film is given in
3. Results and discussion Fig. 5. The sharp peak around 2916 cm1 and 698 cm1 are
assigned to the stretching bands of CAH in styrene repeating units.
3.1. Characterization of the synthesized oil swellable polymer The peak at 1728 cm1 is assigned to the stretching bands of C@O
in ester group of butyl acrylate repeating units. The peak at
As seen from Fig. 4(a), the synthesized polymer dispersion pre- 2848 cm1 is caused by the CAH stretching band in ACH2. More-
sents a typical milky latex. The solid content of the polymer disper- over, the C@C absorption band in the region 1680–1620 cm1 does
sion is 27.9% and the particle size is ca. 180 nm measured by not appeared in the spectrum, which may only originate from the
dynamic laser scattering. The TEM image of the polymer particles polymerizable monomers. All of these information provide clear
is shown in Fig. 4(b). In order to confirm the co-polymerization evidence of the successful co-polymerization of the related
of the various monomers in the synthesis of OSP as described in monomers.
102 Z. Lu et al. / Construction and Building Materials 114 (2016) 98–108
Fig. 3. Schematic diagram of the preparation of the HCP specimen with a single penetrating crack for the flow tests (a) pre-notched cylindrical HCP specimen subjected to
bending break; (b) split HCP specimens; (c) integrated HCP specimen with a single artificial crack.
The incorporation of the long alkyl side chains from the func-
tional monomer is expected to make the obtained polymer highly
oil absorptive and swellable. The saturated oil absorptivity of the
dried OSP film calculated according to Eq. (1) was as high as
700% by weight of the polymer itself, while the water absorptivity
was only 11%. This suggests that the synthesized polymer is highly
hydrophobic.
3.2. Water and Oil absorption of the hardened cement pastes (HCPs)
formed by the volume occupied by water that is not consumed in saturated water absorptivity is determined by the total porosity
the hydration process. Many properties of HCP have direct relation of the capillary pores, the air voids, and the pore connectivity.
to the capillary pores, such as permeability, volume stability and In this paper, the water/oil absorption of dried HCPs was mea-
compressive or flexural strength [32]. Air voids are formed by air sured over 200 h by immersing the HCP samples in a water/oil bath
entrainment during the preparation of cement pastes and usually as shown in Fig. 1. As seen in Fig. 6(a), for the BHCP, the absorption
have larger pore size than 1 lm. It has been well reported that processes of water and oil are characterized by two steps, 1) a fast
the pore structure of HCP is highly related to W/C and hydration absorption within 10 h, and 2) a slow absorption over a longer per-
degree [33]. Specifically, for a given HCP, growth of hydration iod. It has been proposed that the fast absorption process is mainly
degree leads to reduction of the total pore volume, increase of attributed to the flow of the fluid sucked into the large and con-
the gel porosity and decrease of the capillary porosity. For a fully nected pores. The following slower absorption process is deter-
hydrated HCP, higher W/C usually generates larger total porosity, mined by the slow diffusion of the fluid through finer capillary
capillary porosity and higher pore connectivity [33]. pores due to increased friction resistance [37]. Here, we regard
As well understood, because the hydration products of Portland the absorptivity at 200 h as the saturated absorptivity. For BHCP,
cement are highly hydrophilic, a dried HCP is highly water absorp- we find from Fig. 6(a) that the saturated absorptivity for both fluids
tive due to the capillary force. The gel pores usually are occupied is almost equal by mass. However, it should be noted that the den-
by water which is restricted by the capillary force and strong sity of oil is about 0.827 g/cm3 (at 80 °C), remarkably lower than
hydrogen bonds to the C-S-H [34]. The partial dehydration of the that of water (0.972 g/cm3 at 80 °C). We convert the mass absorp-
gel pores in C-S-H was observed at temperature up to 105 °C at tivity (Fig. 6(a)) to the volume absorptivity per unit mass of
atmosphere pressure [35], which leads to decomposition of the cement, as seen in Fig. 6(b). Surprisingly, one can see that the
whole material. Therefore, in a dry HCP (under 105 °C) specimen,
the capillary pores as well as the air voids are the main accessible
volume for water ingression. For a fully hydrated HCP, the
Fig. 6. Mass increase of the HCP samples (a) and the volume of the absorbed fluid
per unit mass of cement (b) during water/oil immersion (in the conversion of (a) to
(b), it is assumed that the mass ratio of a hardened cement paste to the raw cement
is 1.22 due to the acquirement of the non-evaporable water [36]). Fig. 7. Morphology of BHCP (a) and MHCP (b) obtained by SEM.
104 Z. Lu et al. / Construction and Building Materials 114 (2016) 98–108
saturated volume absorptivity of oil is clearly higher than that of absorption in MHCP presents much slower than in BHCP. It is well
water, which suggests that the total accessible volume for oil is known that when a polymer latex is added into cementitious sys-
larger than that for water in BHCP. It should be mentioned that tem, film formation of polymer particles may take place during
for both water and oil absorption tests, the samples were com- hardening due to the consumption of water by both cement hydra-
pletely dried to remove most of the capillary pore water as a start- tion and drying [40,41]. The formed polymer films may obstruct
ing point. In water absorption test, the absorbed water may form a some water diffusion paths through capillary pores, as can be
wetting layer on the surface of pore wall and such water layer observed in Fig. 7 and hence reduces the water absorption speed
could be strongly bound or fixed by the pore wall due to the strong in MHCP.
interaction between water and the highly hydrophilic pore wall. We used the contact angle method to indicate the wetting
This fixed water layer may block some fine pores and hinder, or behavior of the two fluids on the polished surface of HCPs. It was
even stop further migration of water into HCP driven by capillary surprisingly found that the oil presents complete wetting on the
force [38,39]. On the other hand, in the oil absorption test, the for- surface of both BHCP and MHCP despite its hydrophobic feature,
mation of such bound oil layer does not occur because of the while the water exhibits a droplet with a certain contact angle as
absence of the strong interaction between the oil and the pore wall, seen in Fig. 8. The contact angle of water on surface of MHCP is
so that slightly higher oil absorptivity than water absorptivity was clearly bigger than on surface of BHCP. These results indirectly
observed in Fig. 6(b). manifest why the volume absorptivity of oil is larger than that of
On the other hand, it is interestingly seen from Fig. 6(a) that the water on a certain HCP sample and why the water absorption on
water absorption of MHCP is much slower and the saturated water MHCP is slower than that on BHCP, as indicated in Fig. 6(b)
mass absorptivity is notably smaller than that of BHCP in spite of Most importantly, for both Fig. 6(a) and (b), MHCP exhibits a
their equivalent W/C ratios. As BHCP and MHCP were prepared much higher saturated oil absorptivity than its saturated water
with the same W/C ratios and both had been cured for more than absorptivity, and also than the saturated oil absorptivity of BHCP.
28 d, it could be roughly assumed the same total pore volume per The only possibility to explain this is that the incorporated OSP
unit mass of cement for both materials. Thus, when we compare contributes to the increased oil absorption. As described in Sec-
the volume of the absorbed water per unit mass of cement by both tion 3.1, it has been shown that the polymeric film of OSP has a
HCPs as shown in Fig. 6(b), it is seen that the saturated volume strong capability of oil absorption with a saturated oil absorptivity
absorptivity for BHCP and MHCP becomes closer, while the water
Fig. 9. Cumulative pore volume per unit mass of cement measured by MIP.
Fig. 10. Cumulative pore volume per unit mass of cement measured by N2
Fig. 8. Contact angles of DI water on polished surface of BHCP (a) and MHCP (b). adsorption.
Z. Lu et al. / Construction and Building Materials 114 (2016) 98–108 105
of about 700%. Thus, the oil absorption of MHCP may have two
1) No pore size information available and only the total porosity measured
manners, the oil absorption by pores and by OSP films. Assuming
both MHCP and BHCP have the same total accessible pore volume
for oil per unit mass of cement, the increased oil absorptivity of
MHCP should be contributed by the incorporated polymeric film.
adsorption.
The MIP results reveal that the total pore volume per unit mass
Advantages and disadvantages
Disadvantages [48]:
Disadvantages [49]:
1) Easy operation
2) Low accuracy
mass of HCP. Therefore, we convert the pore volume per unit mass
Advantages [47]:
Disadvantages:
Fig. 9. Similarly, it is found that the total pore volume per unit mass
of cement is increased by about 16% due to the addition of OSP.
Because of the same W/C ratio for both BHCP and MHCP, it has
been expected that the total pore volume per unit mass of cement
for MHCP would be equal to or less than that for BHCP (due to the
jamming effect of polymer film in capillary pores). However, this is
not the case as seen in Fig. 9. Recalling that in the MIP measure-
2 nm–100 nm [46]
2 nm–7.5 lm [45]
give more information about the pore structure of the two HCPs.
Effective pores in the measurement
are shown in Fig. 10. From Fig. 10, it is clearly seen that the addi-
tion of OSP decreases the total porosity of MHCP, which is in good
agreement with the test result as discussed in Fig. 6 and with the
MHCP
Total pore volume per
0.037
0.200
unit mass of cement
0.173
0.078
BHCP
that for a given HCP sample, the total pore volumes obtained from
different methods are different due to the different mechanism
involved in the measurements, as summarized in Table 4. For the
Table 4
MIP
blank HCP, the water absorption and the MIP tests give fairly com-
parable results of the total pore volume, whereas the result
106 Z. Lu et al. / Construction and Building Materials 114 (2016) 98–108
obtained from N2 adsorption test presents much lower because of HCP specimen that is completely confined by a steel frame, was
its ineffectiveness for large pores (>100 nm). One the other hand, investigated. The oil flow tests were conducted by regulating the
for the polymer modified HCP, the MIP measurement gives the inlet pressure and the outlet pressure of the oil flow through the
highest total pore volume, which is believed that a part of volume crack was constantly equal to the atmosphere pressure. The pres-
occupied by the polymer films is counted as pore volume by the sure difference between the inlet and the outlet was varied from
MIP measurement because the high pressure liquid mercury may 0.02 to 0.12 MPa.
destroy the polymeric films in the body of HCPs. In this case, water According to the Poiseuille law, the flow behavior of a fluid
absorption and N2 adsorption methods are more relevant. For com- through a thin gap should follow the Eq. (2), at a precondition of
parison of the pore structure of HCPs in the absence and the pres- the laminar flow:
ence of polymer, combination of the three methods provides more
reliable and full-scale results than any single testing method. W 3 DP q
Q ¼L ð2Þ
12lD
3.4. Linear expansion of the hardened cement pastes upon oil
absorption where Q is the mass flow rate (g/s), L is the length of the crack (cm),
W is the crack width (cm), DP is the pressure difference (Pa)
The previous results certainly indicate that the incorporation of between the inlet and the outlet of the fluid flow, q is the fluid
OSP into HCP increases the oil absorptivity of the HCP specimen. It density (g/cm3), D is the depth of the crack (cm), l is the viscosity
is expected that the increased oil absorption by the incorporation of the fluid (Pa s). It is seen from Eq. (2) that the flow rate Q is
of OSP in HCP leads to higher expansion of the HCP samples. The directly proportional to the pressure difference DP and the crack
cylindrical HCP specimens were immersed in an oil bath with a width W. Thus, by measuring the dependence of the flow rate on
regulated temperature constantly at 80 °C and the linear expansion the pressure difference, the crack width could be quantified by
of the cylindrical HCP specimen upon oil absorption was continu- using the Eq. (3). This method is easier and more reliable for quan-
ously measured in function of the immersion time over 120 h, as tification of crack width in cementitious materials, compared to the
given in Fig. 11. popularly used microscope observation, by which the crack width
It is clearly shown that both HCP specimens expand sharply
within the first 10 h of the immersion in an oil bath, and then
the expansion slowly grows with time. The final linear expansion
of MHCP was 1260 lm m1, which is more than doubled compared
to the value of BHCP (525 lm m1). Recalling that the saturated oil
absorptivity for MHCP and BHCP are 16.2% and 12.4% respectively,
the linear expansion of MHCP and BHCP specimens per percent of
oil absorption are 77.5 lm m1 and 42.4 lm m1 respectively. This
clearly indicates that the incorporation of OSP increases the oil
absorption capacity of HCP and more effectively increases the
expansion capability upon oil absorption. The thinkable mecha-
nism for the higher expansion could be the modifications of the
pore structure and the decreased elastic modulus from 17.71 GPa
of BHCP to 8.74 GPa of the MHCP due to the incorporation of OSP.
3.5. Crack healing of the hardened cement pastes upon oil absorption
Fig. 12. Oil flow through the single crack in the BHCP (a) and MHCP (b) which has
Fig. 11. Linear expansion of BHCP and MHCP during immersion in oil bath. been immersed in oil bath for different time.
Z. Lu et al. / Construction and Building Materials 114 (2016) 98–108 107
Fig. 13. Self-healing mechanism of MHCP (a) initial crack, (b) crack healing due to the expansion of the specimen.
could not be quantified due to the non-homogenous distribution of 3) Pore structure measurements show that the addition of OSP
crack width along the length and depth direction. decreases the total porosity of MHCP which confirms that
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi the formation of the polymeric film in MHCP.
3 12Q lD 4) The incorporation of OSP increases the oil absorption capac-
We ¼ ð3Þ
LDPq ity of HCP and more effectively increases the swellability of
HCP upon oil absorption. The self-healing ability of a crack in
Both specimens of BHCP and MHCP were subjected to the oil
the HCP specimens with and without OSP was measured by
flow test to verify self-healing properties of the crack in HCP spec-
oil flow test under different pressures. The oil flow through a
imens, as seen in Fig. 12. It is seen from Fig. 12 that the oil flow
single penetrating crack in the HCP specimen is clearly
through the cracks in both HCP specimens strictly complies the
retarded after a certain period of oil immersion.
Poiseuille law and the slope of the linear curves indicates the width
5) This paper introduces a novel technique to achieve the crack
of the cracks. According to Eq. (3), the crack width can be calcu-
self-healing property triggered by oil leakage by using an oil
lated from the slope in Fig. 12 and the calculated initial crack
swellable polymer to modify the cement paste, which has
widths of the BHCP and MHPC are ca. 15 lm and ca. 25 lm respec-
great potential for improving the durability of the oil wells.
tively. After being stored in the oil bath for varied periods of time,
In addition, it has been well documented that the incorpora-
the specimens were subjected to the oil flow test repeatedly to
tion of polymer materials may greatly improve various prop-
indicate the change of the crack width. From Fig. 12, it is clearly
erties of cementitious materials, such as workability [50],
seen that the width of the crack in BHCP keeps almost unchanged
mechanical properties [51–53], shrinkage and cracking
during the storage in the oil bath for a period of 6 d, while the crack
resistance [54], pore structure, permeability [41,55], durabil-
width in MHPC significantly decreases from ca. 25 lm to ca. 12 lm
ity etc. Therefore, the influences of the addition of OSP poly-
during the period of oil bath storage calculated from Eq. (3). This is
mer on other properties of cement paste, including setting
a firm indication that the crack in MHCP is being healed due to the
behavior, mechanical properties, adhesion to substract,
high volume expansion upon the high oil absorption. The self-
permeability and durability of hardened cement pastes need
healing mechanism is schematically illustrated in Fig. 13. When
to be further studied.
the boundary of the specimen is confined, the volume expansion
of the specimen produces an expansion force in the specimen body
and hence leads to the healing of the cracks. In this way, it can be
expected that the leakage of fluid through the cracks can be Acknowledgment
stopped when longer immersion in oil bath is provided.
The supports from the National Natural Science Foundation of
4. Conclusion China (Grant No. U1262107 and U1301241) are appreciated.
[8] Noda Plywood Mfg. Co., Ltd., Fabrication of hydraulic building boards, JP Patent microscope point-counting procedure, Cem. Concr. Res. 34 (10) (2004)
60028776, Noda Plywood Mfg. Co, Ltd., 1985. 1787–1793.
[9] T. Nishiwaki, Fundamental Study on Development of Intelligent Concrete with [32] J. Zhang, P.Y. Yan, W.Z. Qin, Building Materials (in Chinese), Tsinghua
Self-Healing Capability, Tohoku University, 1997. University Press, Beijing, 2007.
[10] Y. Ohama, K. Demura, Superior properties of epoxy modified mortars and [33] R.A. Cook, K.C. Hover, Mercury porosimetry of hardened cement pastes, Cem.
concretes without hardener, in: Proceedings 3rd Southern African Conference Concr. Res. 29 (1999) 933–943.
on Polymers in Concrete and ICPIC Workshop 1, 1997. [34] D. Snoeck, L.F. Velasco, A. Mignon, S. Van Vlierberghe, P. Dubruel, et al., The
[11] T. Katsuhata, Y. Ohama, K. Demura, Investigation of microcracks self-repair influence of different drying techniques on the water sorption properties of
function of polymer-modified mortars using epoxy resins without hardeners, cement-based materials, Cem. Concr. Res. 64 (2014) 54–62.
in: Proceedings of the 10th International Congress on Polymers in Concrete, [35] N.C. Collier, J.H. Sharp, N.B. Milestone, J. Hill, I.H. Godfrey, The influence of
Honolulu, USA, 2001. water removal techniques on the composition and microstructure of hardened
[12] L. Czarnecki, Sustainable concrete; is nanotechnology the future of concrete cement pastes, Cem. Concr. Res. 38 (2008) 737–744.
polymer composites?, Adv Mater. Res. 687 (2013) 3–11. [36] R.Z. Yuan, Cementitious Material Science (in Chinese), second ed., Wuhan
[13] P. Łukowski, G. Adamczewski, Self-repairing of polymer-cement concrete, Bull. University of Technology Press, Wuhan, 1996.
Pol. Acad. Sci. 61 (1) (2013) 195–200. [37] C.H. Shen, Researches on the Moisture Transport of Cement Based Materials,
[14] K.V. Tittelboom, N.D. Belie, W.D. Muynck, W. Verstraete, Use of bacteria to Wuhan University of Technology, 2007.
repair cracks in concrete, Cem. Concr. Res. 40 (2010) 157–166. [38] S. Diamond, A discussion of the paper ‘‘Effect of drying on cement-based
[15] J. Xu, W. Yao, Multiscale mechanical quantification of self-healing concrete materials pore structure as identified by mercury porosimetry-a comparative
incorporating non-ureolytic bacteria-based healing agent, Cem. Concr. Res. 64 study between oven, vacuum, and freeze-drying’’ by C. Gallé, Cem. Concr. Res.
(2014) 1–10. 33 (2003) 169–170.
[16] A. Hosada, T. Kishi, H. Arita, Y. Takakuwa, Self healing of crack and water [39] C. Gallé, Reply to the discussion by S. Diamond of the paper ‘‘Effect of drying on
permeability of expansive concrete, in: Proceedings of the First International cement-based materials pore structure as identified by mercury intrusion
Conference on Self Healing Materials, Noordwijk, Netherlands, 2007. porosimetry: a comparative study between oven, vacuum and freeze-drying’’,
[17] T. Kishi, T.H. Ahn, A. Hosoda, S. Suzuki, H. Takaoka, Self-healing behavior by Cem. Concr. Res. 33 (2003) 171–172.
cementitious recrystallization of cracked concrete, in: Proceedings of the First [40] M. Gretz, J. Plank, An ESEM investigation of latex film formation in cement
International Conference on Self Healing Materials, Noordwijk, Netherlands, pore solution, Cem. Concr. Res. 41 (2011) 184–190.
2007. [41] Y.R. Zhang, X.M. Kong, Influences of superplasticizer, polymer latexes and
[18] K. Sisomphon, O. Copuroglu, Some characteristics of a self healing mortar asphalt emulsions on the pore structure and impermeability of hardened
incorporating calcium sulfo-aluminate based agent, in: Proceedings of the 2nd cementitious materials, Constr. Build. Mater. 53 (2014) 392–402.
International Conference on Durability of Concrete Structures, Sapporo, Japan, [42] C. Fischer, R. Gaupp, Multi-scale rock surface area quantification-a systematic
2010. method to evaluate the reactive surface area of rocks, Chem. Erde 64 (3) (2004)
[19] T.H. Ahn, T. Kishi, Crack self-healing behaviour of cementitious composites 241–256.
incorporating various mineral admixtures, J. Adv. Concr. Technol. 8 (2010) [43] C.J. Groen, L.A.A. Peffer, J. Pérez-Ramı´rez, Pore size determination in modified
171–186. micro- and mesoporous materials. Pitfalls and limitations in gas adsorption
[20] H.X.D. Lee, H.S. Wong, N.R. Buenfeld, Potential of superabsorbent polymer for data analysis, Microporous Mesoporous Mater. 60 (2003) 1–17.
self-sealing cracks in concrete, Adv. Appl. Ceram. 109 (2010) 296–302. [44] S. Brunauer, P.H. Emmet, E. Teller, Adsorption of gases in multimolecular
[21] J. Lu, Preparation, Characterization and Application of High Oil Absorption layers, J. Am. Chem. Soc. 60 (1938) 309–319.
Resin Latex (dissertation), Zhejiang University, 2006. [45] P. Chindaprasirt, C. Jaturapitakkul, T. Sinsiri, Effect of fly ash fineness on
[22] A.M. Atta, K.F. Arndt, Swelling and network parameters of high oil-absorptive compressive strength and pore size of blended cement paste, Cem. Concr.
network based on 1-octene and isodecyl acrylate copolymers, J. Appl. Polym. Compos. 27 (2005) 425–428.
Sci. 97 (1) (2005) 80–91. [46] Q. Zeng, K.F. Li, T. Fen-chong, D. Patrick, Pore structure characterization of
[23] X.M. Kong, S. Emmerling, J. Pakusch, M. Rueckel, J. Nieberle, Retardation effect cement pastes blended with high-volume fly-ash, Cem. Concr. Res. 42 (1)
of styrene-acrylate copolymer latexes on cement hydration, Cem. Concr. Res. (2012) 194–204.
30 (2015) 23–41. [47] J. Kaufmann, Pore space analysis of cement-based materials by combined
[24] D. Urban, K. Takamura, Polymer Dispersions and Their Industrial Applications, Nitrogen sorption-Wood’s metal impregnation and multi-cycle mercury
Wiley-VCH Verlag GmbH, Weinheim, 2002. intrusion, Cem. Concr. Compos. 32 (2010) 514–522.
[25] A. Beeldens, D. Van Gemert, H. Schorn, Y. Ohama, L. Czarnecki, From [48] C.A. Léon y Léon, New perspectives in mercury porosimetry, Adv. Colloid
microstructure to macrostructure: an integrated model of structure Interface Sci. 76–77 (1998) 341–372.
formation in polymer modified concrete, Mater. Struct. 38 (6) (2005) 601–607. [49] H.Y. Ma, Z.J. Li, Realistic pore structure of Portland cement paste: experimental
[26] D.W. Fowler, Polymers in concrete: a vision for the 21st century, Cem. Concr. study and numerical simulation, Comput. Concr. 11 (4) (2013) 317–336.
Compos. 21 (1999) 449–452. [50] X.M. Kong, H. Liu, D.M. Wang, The influence of silanes on hydration and strength
[27] R. VocÏka, C. GalleÂ, M. Dubois, P. Lovera, Mercury intrusion porosimetry and development of cementitious systems, Cem. Concr. Res. 67 (2015) 168–178.
hierarchical structure of cement pastes theory and experiment, Cem. Concr. [51] Q. Wang, P.Y. Yan, X.M. Kong, Compressive strength development and
Res. 30 (2000) 521–527. microstructure of cement-asphalt mortar, J. Wuhan Univ. Technol. 26 (5)
[28] F. Puertas, H. Santos, M. Palacios, S. Martínez-Ramírez, Polycarboxylate (2011) 998–1003.
superplasticiser admixtures: effect on hydration, microstructure and [52] Y.L. Liu, X.M. Kong, Y.R. Zhang, Static and dynamic mechanical properties of
rheological behaviour in cement pastes, Adv. Cem. Res. 17 (2005) 77–89. cement-asphalt composites, J. Mater. Civ. Eng. 25 (2013) 1489–1497.
[29] S. Diamond, Mercury porosimetry – an inappropriate method for the [53] H.L. Yang, X.M. Kong, Y.R. Zhang, Mechanical properties of polymer-modified
measurement of pore size distributions in cement-based materials, Cem. silica aerogels dried under ambient pressure, J. Non-Cryst. Solids 357 (2011)
Concr. Res. 30 (2000) 1517–1525. 3447–3453.
[30] D.A. Silva, V.M. John, J.L.D. Ribeiro, H.R. Roman, Pore size distribution of [54] X.M. Kong, Z.L. Zhang, Z.C. Lu, Effect of pre-soaked superabsorbent polymer on
hydrated cement pastes modified with polymers, Cem. Concr. Res. 31 (2001) shrinkage of high-strength concrete, Mater. Struct. 48 (9) (2015) 2741–2758.
1177–1184. [55] X.M. Kong, C.H. Wu, Y.R. Zhang, X.L. Li, Polymer modified mortar with gradient
[31] X. Feng, E.J. Garboczi, D.P. Bentz, P.E. Stutzman, T.O. Mason, Estimation of the polymer distribution: preparation, permeability, and mechanical behavior,
degree of hydration of blended cement pastes by a scanning electron Constr. Build. Mater. 38 (1) (2013) 195–203.
Note: This service is not intended for secure transactions such as banking, social media, email, or purchasing. Use at your own risk. We assume no liability whatsoever for broken pages.
Alternative Proxies: