Documente Academic
Documente Profesional
Documente Cultură
Abstract: In many industrial processes, the pipeline systems are lined with a protective layer of cement mortar. In petroleum wells, cement slurry is placed in a wellbore
to be hardened into an impermeable mass that seals the annulus from fluid flow and
protects the casing from corrosion for the life of the well. When uniform linings of
neat cement fail in tension, one or more large cracks are formed and the pressurizing
fluid or mud easily flows through the cracks. The necessity to check the damaging
effect of plastic shrinkage in cement mortar, and thus the formation of cracks, has
called for further studies in this topic. In the past, the most common research topic
has been in the areas of polymer fibers that are expensive and environmentally unacceptable. In the quest of pursuing technologies that are environmentally friendly,
inexpensive, and innovative, this paper suggests the use of human hair, a waste material, in order to replace polymer fibers. Hair waste has been used as a new natural
fiber to reinforce mortar and cement and improve their impermeability. The investigation reported herein concerns the effects of human hair fibers on the reduction of
shrinkage cracking of mortar. The influence of mix proportions on the plastic shrinkage of human hair fiber reinforced mortar has been studied. The approach selected
in this study is based on the factorial design of experiments, in which the considered
parameters are cement/sand ratio, water/cement ratio, and human hair fibers content.
The results show that human hair fibers are effective in reducing the plastic shrinkage
cracks area of mortar by a remarkable percentage up to 92%.
Keywords: shrinkage cracking, mortar, human hair waste, petroleum well cementing
Address correspondence to M. M. Al-Darbi, Killam & NSERC Post-Doctoral Fellow, Department of Materials Engineering, The University of British Columbia, 3096350 Stores Road, Vancouver, B.C. V6T 1Z4, Canada. E-mail: aldarbi@interchange.
ubc.ca
1267
1268
M. M. Al-Darbi et al.
INTRODUCTION
Cement mortar lining is a process by which metal pipelines are coated internally to protect their structures. There are several advantages of the cement
mortar lining of pipes compared to other pipeline protection methods. Some
of those advantages are: preventing pipe leakage, protecting the inner surface
of the pipe against corrosion by forming an alkaline environment in contact
with the pipe material where corrosion of the steel is inhibited, decreasing
the pumping energy consumption by providing a smooth interior surface with
a high flow coefficient, and as a result, reducing pipe maintenance.
The objective of cementing the annulus, which is present between the
casing and the formation, is to provide zonal isolation of the formations
that have been penetrated by the wellbore. It has been known for decades
that cemented annuli in the wellbore are prone to leaks and hence can lead
to corrosion (Gollapudi, 1993; Talabani and Islam, 2000). However, no fluid
communication should develop during the life of the well among these various
formations, whether they are saturated with water, oil, or gas, and the surface
(Thiercelin et al., 1998; Nowak and Patout, 1997). When uniform linings of
neat cement fail in tension, one or more large cracks are formed and the
pressurizing fluid or mud easily flows through them. When fiber containing
cement fails in tension, they usually form large numbers of small cracks. The
cement matrix fails first by forming microcracks, and then the fibers take over
the loading. The fiber-laced cracks give a high resistance to fluid leak off.
When fiber cement samples are subjected to high impact loads, the cement
matrix shatters, but the fibers hold the broken matrix together (Stewart et al.,
1997).
In addition, fibers reduce the plastic shrinkage of cement. In the setting
process, cement slurries shrinkage and that causes the cement hydrostatic
pressure to drop. The hydrostatic pressure is important, as gas starts to flow
into the cement when the pressure of the cement column falls below that
of a gas-bearing formation. After the gas has entered the pore system of the
cement, the gas inside may overcome the tensile strength of the cement structure, break the cement matrix, and migrate through the microfractures. A low
shrinkage rate is preferable because the decline in the resulting hydrostatic
pressure will be slower than that for slurry with a higher shrinkage rate. Slow
shrinkage has two advantages: (a) the pressure equilibrium between formation and slurry columns can be reached, and (b) the driving force behind the
flow of pore fluid into the cement will be lower. Both factors should reduce
the risk of early time gas migration (Backe et al., 2001, 1999; Sabins and
Wiggins, 1997).
The cement sheath integrity is important for safe and economical operation of gas storage, geothermal and producing wells. Loss of cement integrity
can cause the following serious events: loss of gas reserves, unsafe operations,
premature water of gas cap production, extra costs because of unplanned remedial operations, and well shutdown (Bybee, 2002).
1269
Plastic shrinkage cracks are random cracks that occur in the exposed
surface of fresh mortar during the first few hours after the mortar is placed,
while the mortar is still plastic and before attaining any significant strength
(Shaeles and Hover, 1988; Samman et al., 1996). As drying starts, the mortar
near the surface dries and shrinks faster than the inner mortar, causing tensile
stress and possible cracks. Plastic shrinkage cracking is usually associated
with hot-weather concreting; however, it can occur within ambient conditions
that produce rapid evaporation of moisture from the mortar surface (Kosmatka
et al., 1995).
Cement and mortar products are notable for their weakness in tension
and for their lack of toughness, which gives risk to frequent cracking under
impact loads, thermal shocks, or dimensional changes due to humidity variation. Fibers have been used for decades to overcome such deficiencies and
to improve the impermeability and minimize shrinkage, which are essential
requirement properties of concrete besides its strength. The three main types
of fibers that may be used as reinforcement for concrete are steel, glass, and
organic (natural and synthetic) fibers. As far as natural fibers are concerned,
animal and vegetable fibers (i.e., wood-cellulose, sisal, bast, coconut, and
bagasse) are all being used in various sheet materials (Padron and Zollo,
1990; Krenchel and Jensen, 1980; Majumdar and Nurse, 1978). The drawbacks of using these natural materials are: the high water absorption which
must be allowed for in the mixing process, the biological deterioration of the
fiber if not treated, and the strength loss that may occur in alkaline environments. Additional processing may be required, as in the case of bagasse, to
remove the sugar from the fiber (Soroushian and Ravanbakhsh, 1998; Cook,
1980).
In this paper, human hair waste is introduced as a new mortar-reinforcing
material. It would be highly pertinent to talk about the hair morphology, mechanical, chemical, physical, and electrical characteristics, to elucidate and
provide the reader with the unique characteristics of hair that make it potentially usable as a composite in cement matrix. Human hair consists of five
definite morphological components: cuticle, cortex, medulla, melanin granules, and cell membrane complex, each distinct in morphology and chemical
composition. Human hair consists of approximately 80% protein, 15% water,
and 5% lipids. The water content of hair varies directly with the ambient
relative humidity (Potsch, 1995).
Regarding the hair mechanical properties, the load required to obtain
breakage of a natural and healthy hair varies between 50 and 100 g. The
average healthy human head, which contains approximately 120,000 hairs,
may handle 12 metric tons (Katz and Chatt, 1988). For an average hair, the
distribution point corresponds to a load of 12 kg/mm2 , and this exceeds that of
aluminum. The unusual hair strength is caused by the keratin, which is a type
of protein found in the hair cortex. The long keratin molecules in the cortex
are compressed to form a regular structure, which is not only strong but also
flexible. Keratin is unique in that its chains contain high concentrations of a
1270
M. M. Al-Darbi et al.
particular amino acid called cystine. Every cystine unit contains two cystine
amino acids in different chains that have come to lie near to each other and
are linked together by two sulfur atoms, forming a very strong chemical bond
known as a disulphide linkage. Many disulphide bonds form down the length
of the keratin chains, joining them together like the rungs of a ladder (Gray,
2000).
Hair has a high frictional coefficient, higher than that of vegetable or
synthetic fibers. The high frictional coefficient of hair is attributed to its
special surface structure (presence of scales), as can be seen in Figure 1
(Katz and Chatt, 1988). Hair is permeable to water in liquid form as well as
to water vapor. After sufficient contact, hair keratin can absorb water up to
35 or 40% of its weight. The absorbed water is partially linked to the keratin
protein by hydrogen bonds, but it can also exist in the free form. When water
is absorbed by keratin the hair diameter can increase by 1520% and its
length can increase only by 0.52%. The water absorption and subsequent
swelling depends mainly on the pH level. Swelling is limited if the pH is
low, and greatly enhanced if the pH is high (Zviak, 1986).
It was found recently that human hair is a good absorbent material. For
instance, in 1998 the US EPAs (Environmental Protection Agency) Oil Spill
Program Internal Journal published an astonishing report on hair as a good
absorbent material for crude oil spill clean up. It was reported therein that
NASA (the National Aeronautical and Space Administration) was testing an
unusual absorbent material for oil spill clean up with human hair, after a
1271
EXPERIMENTAL
Materials and Methods
Portland cement (type I) was selected for the entire program of work. The
chemical properties of this cement are listed in Table 1. Silica sand with
grains proportioned greater than 0.125 mm, waste human-hair pieces with
length greater than 2 cm, and tap water were also used.
wt%
SiO2
Fe2 O3
TiO2
P 2 O5
Mn2 O3
Al2 O3
CaO
MgO
Na2 O
K2 O
SO3
Loss on ignition
Free lime
21.6
1.84
0.31
0.35
0.15
4.3065
65.28
1.18
0.15
0.42
3.06
1.04
1.94
1272
M. M. Al-Darbi et al.
c/s
w/c
Plain
Fibrous
0
0.4
1:1
1:1
0.42
0.42
1273
The same procedure was followed for the factorial design study after an
encouraging breakthrough. The upper and lower values for each variable are
listed in Table 3.
Lower value
Upper value
fh (vol. %)
c/s
w/c
0
1:1
0.38
0.4
1:1.5
0.50
1274
M. M. Al-Darbi et al.
type and strength between the matrix and the fiber. It can also be observed
from Figure 3 that the sheen disappeared from the surface of the plain mortar
slab, whereas the sheen remained in the fibrous mortar slab. Generally, the
disappearance of the sheen from the surface of concrete indicates the time
when the rate of evaporation exceeds the rate at which bleeding water rises
1275
Figure 4. Reduction in the shrinkage cracks area in the fibrous mortar compared to
plain mortar slabs.
to the surface (Shah et al., 1998). This shows that the rate of evaporation is
greater than the rate of rise of the bleeding water in the plain mortar, but in the
fibrous mortar the bleeding rate is retarded by the hair fibers. From that it can
be concluded that the quantity of surface water is significantly reduced by the
addition of fibers. Also, fibers seem to cause a reduction in consolidation, thus
eliminating the damaging capillary bleed channels and causing an increase
in intergranular pressure in the plastic mortar. It was also observed that the
first portion of shrinkage developed slowly. The evaporation of water caused
an increase in the capillary pores (vacuum) pressure and decrease in volume,
which in turn caused contraction (Soroushian et al., 1995).
The high reduction of shrinkage cracks can also be a result of effective
diameter of the hair fibers, which varies between 57 to 120 m (Gray, 2000),
and is comparable to the cement particles size. This promotes a close and
very effective packing and development of a dense bulk and interface microstructure in the matrix (Walton and Majumdar, 1978). The relatively high
surface area and the close spacing of hair fibers make them quite effective in
the stabilization and suppression of microcracks in the mortar samples.
The length of the fibers and their distribution affect the reduction of
cracks in two ways. If the fibers are relatively long and far apart, they will
have no ability to arrest micro-cracks, but they can arrest the propagation of
macro-cracks and substantially improve the toughness of the composite. This
is probably due to the fact that matrix cracking first occurs at the micro level.
However, small fibers could bridge even the micro-cracks.
The high aspect ratio of the fiber (length/diameter) may be accountable
for high reduction of cracks. The hair fibers used in this study have small
diameter range. The small diameter gives the fiber less surface area and,
consequently, fewer flaws that might propagate during cracking. The aspect
ratio for the hair fiber used with a length of 2 cm ranges between 160 and 350.
1276
M. M. Al-Darbi et al.
For the same length (2 cm) and a diameter range between 0.20 to 0.38 mm,
the aspect ratio for steel fibers would range between 26 and 80, and for
polypropylene with a diameter of 0.38 mm, the aspect ratio would be 53
(Lee, 1992). Another explanation would be that the hair fibers increase the
amount of large pores in cement paste. These groups of pores are probably
attributed to the interfacial zone between the fiber and the cement paste. The
formation of large pores may reduce capillary pressure in the cement paste,
thus reducing plastic shrinkage cracking.
Concerning the effect of the hair morphology on the reduction of shrinkage cracks, hair has a high friction coefficient, higher than that of vegetable
or synthetic fibers, as was shown in Figure 1. These high frictional values
could be due to its special surface structure and the presence of scales and
their orientations. The high friction increases the shear and friction forces
between the fibers and the cement matrix. As the matrix shrinks, a shear
stress along the fiber and matrix interface develops. The fiber is subjected
to tension, and the matrix is subjected to compression. The shrinkage of the
matrix in any direction can be considered to be restrained by an aligned fiber
of effective length parallel to the direction of the shrinkage strain (Zhang and
Li, 2001).
Another explanation could be the behavior of hair when in contact with
fluids. Hair is permeable to water in liquid form as well to water vapor.
It was reported that after sufficient contact, hair keratin could absorb up to
35 or 40% of its weight in water (Katz, 1988). This apparently explains
why hair could be a good sorbent material in the oil spill clean up. With
absorption of water as high as the above values, the hair fibers are capable of
retaining a substantial amount of evaporation from the mortar surface and this
reduces the shrinkage cracks. The water absorption and subsequent swelling
depends mainly on the pH level. Swelling is limited if the pH is low (acidic),
and greatly enhanced if the pH is high (alkaline). Since mortar has a high
pH value, approximately 1112, the swelling of mortar is high and, as a
consequence, this results in a high shrinkage reduction.
1277
This type of experimental design with three variables requires eight experiments. These sets of experiments are represented as the number of spheres
at the corners of the cube (Box-Hunter Cube). From the possible combinations, 23 , and according to the factorial design method, the sets of experimental data sets were generated and presented in Figure 5.
Using the Box-Hunter statistical method, the main effect of each variable
was determined, then the interaction effect between two factors was evaluated.
Finally, the effect of the three variables together was calculated. The main
effect, which is the average effect of a variable over all conditions of the
other variables, is calculated from the data, Yi , obtained at the eight corners
i = 18 as follows:
Main effect = Y + Y
(1)
where Y + is the average response for the upper level of the variable, and Y
is the average response for the lower level.
fh effect =
Y2 + Y4 + Y6 + Y8 Y1 + Y3 + Y5 + Y7
4
4
(2)
c/s effect =
Y3 + Y4 + Y7 + Y8 Y1 + Y2 + Y5 + Y6
4
4
(3)
w/c effect =
Y5 + Y6 + Y7 + Y8 Y1 + Y2 + Y3 + Y4
4
4
(4)
1278
M. M. Al-Darbi et al.
Y 1 + Y 4 + Y 5 + Y 8 Y2 + Y 3 + Y 6 + Y 7
4
4
(5)
fh w/s effect =
Y 1 + Y3 + Y6 + Y8 Y 2 + Y4 + Y5 + Y7
4
4
(6)
Y1 + Y2 + Y7 + Y8 Y3 + Y4 + Y5 + Y6
4
4
(7)
Finally, the interaction between all the three parameters is given by:
fh c/s w/c effect =
Y 5 + Y 6 + Y7 + Y 8 Y 1 + Y 2 + Y 3 + Y 4
4
4
(8)
Figure 6 shows the shrinkage area vs. the c/s and w/c ratios for both
plain mortar (fh = 0 vol%) and fibrous mortar (fh = 0.4 vol%). It can be
observed that the highest crack areas were in the mixture with w/c = 0.47,
c/s = 1, and fh = 0 vol%. This result was expected, since the higher level
of w/c ratio was used with lower level of c/s ratio, and no fiber was added.
This is in contrast to the mix with c/s ratio = 0.47, w/c ratio = 1:1.5, and
Figure 6. The percentage of shrinkage cracks areas vs. the c/s and w/c for plain
mortar (fh = 0 vol%) and fibrous mortar (fh = 0.4 vol%).
1279
hair fiber volume fraction = 0.4%. It was observed that the least number of
cracks was formed with these combinations. Again, this met the expectations,
since the lower level of w/c ratio was used with the higher level of c/s ratio.
Table 4 represents the results of the factorial design analysis. The main
effect of a factor should be individually interpreted only if there is no evidence that the factor interacts with other factors. When there is evidence that
one or more of such interactions are there, the interacting factors should be
considered jointly. The negative sign of the values shows a reduction effect
on shrinkage area. For example, the effect of c/s ratio is to reduce the shrinkage by 11 units, and this is the highest main effect in reducing shrinkage
cracks area. That was expected (Popovics, 1982) since sand is more or less
an inert material that greatly reduces the amount of cement in the mixture,
thus decreasing the creep and shrinkage of the concrete. Besides that, the
quartz sand which is the type of aggregate used in the entire program is a
low adsorption aggregate. Low adsorption aggregates have a low shrinkage
property and vice versa (Kosmatka et al., 1995).
It can be observed from Table 4 that there is a large c/s ratio and fh
volume fraction effects, 11 and 9 units, respectively. From that we conclude
that the hair fiber effect in the reduction of shrinkage cracks is comparable
to the effect of c/s ratio. But, since cement ratio interacts well with hair fiber
(the fh c/s interactions is 8 units), we make no statement about the effect
of hair fiber or c/s ratio alone.
The main effect of w/c ratio shows an increase in shrinkage by only
1.53 units, and there is no evidence of any interactions involving w/c ratio.
That was not expected since the most important controllable factor affecting shrinkage is the amount of water per unit volume of concrete. Many
Table 4. Estimated effects from the
23 factorial design
Main effects and
interactions (102 )
Parameter
Main effect
fh
c/s
w/c
9.21
11.01
1.53
Two factor interactions
fh w/c
fh c/s
c/s w/c
1.67
8.66
0.927
6.00
1280
M. M. Al-Darbi et al.
researchers have reported the importance of the effect of w/c ratio. For instance, the work performed at the Massachusetts Institute of Technology
showed that for each 1% increase in mixing water, concrete shrinkage increased by 2% (Kosmatka et al., 1995). However, the result as presented here
shows the opposite. It is my belief that this could be due to the fact that
this study is the first of its kind to use human hair as a cementing fiber, thus
further research is needed to study this effect.
Thus, we draw the following tentative conclusions from the factorial
design: the effect of w/c ratio is to increase the shrinkage by 1.5 units, and
this is irrespective of the tested levels of the other factors. The effects of hairfiber volume fraction and c/s ratio cannot be interpreted separately because
of the large fh c/s interaction. The shrinkage cracks increase as the hair
fiber volume fraction and c/s ratio decrease.
CONCLUSIONS
A novel method for reducing the shrinkage of a cement mortar is developed.
This shrinkage arises from unequal dissipation of water from the cement
mortar and the rising of bleeding water to the mortar surface. The proposed
process involves the use of human hair, a waste material, to improve the
quality of cement, as applied in petroleum wells or in lining pipelines. From
this study, the following tentative inferences can be drawn. The hair reduced
the shrinkage cracks area by 92%. The hair-fiber volume fraction effect in
reducing the shrinkage cracks is close to the established cement/sand ratio
effect. The experimental results indicate that the water/cement ratio does not
have significant impact on shrinkage whenever hair fibers were added to the
cement sample. This may be responsible for the reduction in shrinkage with
hair fibers. The results of this study also show that the commercial application
of human hair waste fiber on shrinkage reduction is feasible, such as in cement
coated pipelines and cementing hydrocarbon wells.
ACKNOWLEDGMENTS
The authors would like to thank the Natural Sciences and Engineering Research Council of Canada (NSERC), Petroleum Research Atlantic Canada,
Atlantic Innovation Fund, and Shell Canada for their funding. Special thanks
from the first author to NSERC and Killam Foundation for their prestigious
scholarships and the generous financial support.
REFERENCES
Ali, M. A., Majumdar, A. J., and Singh, B. (1978). Properties of glass fiber
cement: The effect of fiber length and content (CP 94/75). In Fiber
Reinforced Materials. Lancaster: The Construction Press.
1281
1282
M. M. Al-Darbi et al.
Popovics, S. (1982). Fundamentals of Portland Cement Concrete: A Quantitative Approach. New York: A Wiley-Interscience Publication.
Potsch, L. (1995). In Physiology and Ultra-Structure of Human Hair. Proceedings of the International Conference and Workshop for Hair Analysis
in Forensic Toxicology, Abu-Dhabi, United Arab Emirates, November.
Sabins, F., and Wiggins, M. L. (1997). Parametric study of gas entry into
cemented wellbores. SPE Drilling and Completion 12(3):180187.
Samman, T. A., Mirza, W. H., and Wafa, F. F. (1996). Plastic shrinkage
cracking of normal and high-strength concrete: A comparative study.
ACI Materials Journal 93(1):3640.
Shaeles, C. A., and Hover, K. C. (1988). Influence of mix proportions and
construction operations on plastic shrinkage cracking in thin slabs. ACI
Materials Journal 85(6):495504.
Shah, S. P., Ouyang, C., Markunte, S., Yang, W., and Becq-Giradon, E.
(1998). A method to predict shrinkage cracking of concrete. ACI Materials Journal 95(4):339347.
Soroushian, P., Mirza, F., and Alhozaimy, A. (1995). Plastic shrinkage cracking of polypropylene fiber reinforced concrete. ACI Material Journal
92(5):553558.
Soroushian, P., and Ravanbakhsh, S. (1998). Control of plastic shrinkage
cracking with specialty cellulose fibers. ACI Materials Journal 95(4):
429435.
Stewart, R. B., Gill, D. S., Lohbeck, W. C. M., and Baaijens, M. N. (1997).
An expandable-slotted-tubing, fiber-cement wellbore-lining system. SPE
Drilling and Completion 12(3):163167.
Talabani, S., and Islam, M. R. (2000). A new cementing technology for
controlling downhole corrosion. Journal of Petroleum Science and Engineering 26(14):4348.
Tan, T. C., Chia, C. K., and Teo, C. K. (1985). Biosorbent removal of metal
by human hair. Water Resources Journal 19:157159.
Thiercelin, M. J., Dargaud, B., Baret, J. F., and Rodriguez, W. J. (1998).
Cement design based on cement mechanical response. SPE Drilling and
Completion 13(4):266274.
Walton, P. L., and Majumdar, A. J. (1978). Cement-based composites with
mixtures of different types of fiber (CP 81/75). In Fiber Reinforced Materials. Lancaster: The Construction Press.
Waters, G. A., and Wray, B. L. (1995). Narrow annulus cementing in the
deep Anadarko basin. SPE Drilling and Completion 10(1):5360.
Zhang, J., and Li, V. C. (2001). Influences of fibers on drying shrinkage
of fiber-reinforced cementitious composite. Journal of Engineering Mechanics 127(1):3744.
Zviak, C. (1986). The Molecular and Structural Biology of Hair. New York:
The New York Academy of Sciences.