-
328DOI DOI 10.1007/s12182-011-0149-6
Wang Chengwen1 , Wang Ruihe1, Cheng Rongchao2 and Chen Erding31
School of Petroleum Engineering, China University of Petroleum,
Qingdao, Shandong 266555, China2 Department of Drilling Strategy
and Planning, CNPC Drilling Research Institute, Beijing 100195,
China3 Drilling Engineering and Technology Corp., Shengli Petroleum
Administration Bureau, Dongying, Shandong 257064, China
China University of Petroleum (Beijing) and Springer-Verlag
Berlin Heidelberg 2011
Abstract: To address present concerns about thickening time and
high early-strength in deepwater cementing at low temperatures when
using conventional accelerators, a new type of set-accelerating
admixture comprising of lithium chloride, aluminium hydroxide and
alkaline metal chlorides, named as LS-A, was studied in this paper.
Mechanism analysis and performance tests show that the accelerator
LS-A accelerated the hydration of tri- and dicalcium silicates (C3S
and C2S) at low-temperatures by speeding up the breakdown of the
protective hydration fi lm and shortening the hydration induction
period. Therefore, LS-A could shorten the low-temperature
thickening time and the transition time of critical gel strength
from 48 to 240 Pa of the Class-G cement slurry, and improve the
early compressive strength of set cement at low-temperatures. It
exhibited better performance than calcium chloride and had no
effect on the type of hydration products, which remain the same as
those of neat Class-G cement, i.e. the calcium silicate gel,
Ca(OH)2 crystals and a small amount of ettringite AFt crystals.
LS-A provides an effective way to guarantee the safety of cementing
operations, and to solve the problems of low temperature and
shallow water/gas fl owing faced in deepwater cementing.
Key words: Deepwater cementing, accelerator, lithium chloride,
Class-G cement, mechanism
Mechanism and performance of a lithium chloride accelerator
*Corresponding author. email: [email protected] October 2,
2010
Pet.Sci.(2011)8:328-334
1 IntroductionApproximately 57 billion barrels of oil equivalent
(BBOE)
hydrocarbons has been discovered in deepwater, and the
yet-to-be-discovered resources are estimated to be 85-100 BBOE
(Pettingill and Weimei, 2002). Many oil companies are showing an
increasing interest in exploration and production of the abundant
hydrocarbon resources in deepwater. Successful deepwater cementing
plays a critical role in assuring the efficient, cost-effective,
and safe development of deepwater hydrocarbon resources. In
deepwater wells, the seabed temperature is usually lower than 4 C,
and the circulating temperature in surface cementing typically
ranges from 10 to 15 C (Rae and Lullo, 2004; Ravi et al, 1999). The
temperature is verified to be a key factor contributing to the
hydration rate of cement slurry. Low temperature will dramatically
reduce the cement hydration rate, which can cause much longer
thickening times, slow development of compressive strength of the
set cement and insuffi cient shear stress of the annular cement
sheath to support the casing within a short period. This will
inevitably prolong the waiting-on-cement (WOC) time and increase
well construction cost
(Wang et al, 2008; 2009). Therefore, favorable early strength is
required for
deepwater cement slurry at low temperatures to achieve a
preferable WOC time of less than 24 hours. An accelerator is an oil
well cement additive used to shorten the thickening time and
enhance the early compressive strength. Concerning the potential
issues in deepwater low-temperature cementing, calcium chloride is
commonly used to improve the early strength of the set cement, in
reference to relevant cementing experience in onshore oilfi elds
(Griffi th, 1996). It is of great importance to develop a new
accelerator used for deepwater cementing at low temperatures. Many
proprietary accelerators have been reported since 2002, such as
polyhydroxyamine compounds (Reddy and Fitzgerald, 2002), a
set-accelerating admixture comprising an alkaline and alkaline
earth metal nitrite (Maberry et al, 2005), and (CaO)m(SiO2)nxH2O
(Ravi et al, 2005). These accelerators have, to an extent,
addressed some deepwater cementing challenges, but their extended
application on a large scale is restricted due to their limited
performance.
Alkali metal chlorides are a conventional type of
set-accelerating materials, especially sodium and potash chlorides.
Lithium chloride is also an alkali metal chloride. In aqueous
solution, lithium cations with a small radius are easily hydrated
due to their strong polarization and the
-
329
hydrated ions have a large hydration radius. This will affect
the cement hydration and thus the compressive strength of set
cement. Since 1951, almost all related studies and reports have
been focused on the suppression of alkali-silica reaction by
lithium chloride (Yu et al, 2009). Brothers and Palmer (2004)
investigated the accelerating effect of lithium chloride on the
setting of cement slurry at low-temperatures. To date, the
acceleration mechanism of lithium chloride and its influence on the
performance of the cement slurry, however, have not yet been
reported in the literature. In this paper, the authors report the
infl uence of lithium chloride on hydration of the Class-G cement
at low temperatures. Based on the acceleration mechanism of lithium
chloride, a set-accelerating admixture comprising lithium chloride,
aluminium hydroxide and alkaline metal chlorides, named as LS-A,
was prepared, and its infl uence on the performance of Class-G
cement at low-temperatures was discussed.
2 Experimental
2.1 MaterialsA set-accelerating admixture comprising of
lithium
chloride, aluminium hydroxide and alkaline metal chlorides,
named as LS-A, was prepared in the laboratory and its element
composition is as follows: Li 14.17 %, Cl 47.73 %, O 27.34 %, Al
4.70 % and Na 4.01 %. Class G HSR cement was provided by Shengli
Huanghe Cementing Corp. and its chemical and mineral composition is
shown in Table 1. Calcium chloride, absolute ethyl alcohol, and
acetone (analytic pure) were provided by Shanghai Branch of China
Pharmaceutical Group.
Table1 Chemical and mineral composition of Class-G cement
used
Chemical composition, wt% Mineralogical composition, wt%
SiO2 Al2O3 Fe2O3 CaO MgO SO3 C3S C2S C4AF C3A
24.76 2.89 2.63 65.08 0.83 1.25 53.70 30.46 8.0 2.8
2.2 MethodsAbsolute ethyl alcohol was used to terminate
hydration
of selected representative samples of set cement cured at 4 C.
After being milled into 75 m powders, the samples were dried under
vacuum at 40 C to a constant weight. An Xpert PRO MPD X-ray
diffractometer (Panalytical Co. Ltd., Netherlands ) was used for
phase identifi cation of hydration products. Test parameters were:
Cu K radiation, 40 kV and 40 mA, scanning range 2=5-70. The dried
cement powders were dispersed on a copper stub using a conductive
adhesive and gold-coated in vacuum, and hydration products of
Class-G cement produced at low-temperatures were observed with a
field emission scanning electron microscope JSM7600F produced by
JEOL Corp., Japan at an acceleration voltage of 3.0 kV.
Cement slurries were prepared following API 10B-3-2004
standards, with a water-to-cement ratio of 0.44. All thickening
times of cement slurries at low temperatures were performed on an
OWC-2000A pressurized consistometer equipped with a circulation
system to control temperature (produced by
Shenyang Petroleum Instrument Research Institute). Static gel
strength development was measured with a static gel strength
analyzer (Model 5265U with UCA functionality, American Chandler
Corp.) for all cement slurries, and all data acquisition was
performed with a Chandler 5270 DACS system. The compressive
strength of set cement was tested after the prepared cement slurry
was cured in a copper mold (50mm50mm50mm) in a self-made
multifunctional curing pot SL-B at 4, 10 and 20 C,
respectively.
3 Acceleration mechanism of LS-A
3.1 Effect of LS-A on low-temperature hydration of oil well
cement
Five successive stages of silicate cement hydration can be defi
ned (Nelson, 1990) (I) preinduction period, (II) induction period,
(III) acceleration period, (IV) deceleration period, and (V)
diffusion period. Once the cement particles contact water, cement
hydration will take place, which means the beginning of the
preinduction period. After several minutes, calcium silicate
hydrate (CSH) gel is formed and precipitates on the surface of
cement grains, and the gel gradually covers each grain, acting as a
protective fi lm. The generation of the protective film indicates
the end of the preinduction period and the beginning of the
induction period (Kjellsen and Justnes, 2004; Kjellsen and
Lagerblad, 2007). Fig. 1 shows SEM images of cement slurry hydrated
at 4 C for different times. The cement slurry was made of tap water
and Class-G cement at a water/solid ratio of 0.44, with an addition
of 3% LS-A. In Fig. 1(a), a CSH film was observed surrounding the
cement particles after seven hours of hydration. As the hydration
went on, the concentration of calcium ions outside the fi lm and
the concentration of silicate ions inside the fi lm continuously
increased, which resulted in the protective fi lm gradually
breaking due to high osmotic pressure (Fig. 1(b) and Fig. 1(c)).
The breakdown of the protective film marks the beginning of the
hydration acceleration period (Nelson, 1990). During this period,
exposed tricalcium silicate and dicalcium silicate particles would
hydrate. A great amount of CSH gel, produced in a Silicate Garden
mode (Double and Hellawell, 1976), cross-linked to form a net
structure (Fig. 1(d)), thus causing the cement slurry to harden.
The SEM microstructure analysis indicates the LS-A may accelerate
the breakdown of the hydration film, resulting in its rapid
disappearance within 2 hours. The protective film of the
conventional Portland cement, however, can exist as long as 3-4
hours at ambient temperatures (Lu et al, 2004). This shows that
LS-A accelerated the breakdown of the protective fi lm and
contributed to shortening the induction period.
To further investigate the effect of LS-A, the hydration
processes of the class-G cement with different accelerators were
tested at 20 C. Fig. 2 indicates that the addition of LS-A did not
change the hydration process of the Class-G cement, but
significantly speeded the cement hydration and shortened the
induction period. This means that the addition of LS-A in cement
slurry reduced or practically eliminated the induction period,
accelerating signifi cantly the hydration reaction in the
acceleration period. Therefore, LS-A would accelerate Class-G
cement hydration at low temperatures.
Pet.Sci.(2011)8:328-334
-
330 Pet.Sci.(2011)8:328-334
-
331
Fig. 3 XRD patterns of cement samples after 12 hours of
hydration at 4 C
10 20 30 40 50S1
2
S2
S1: Dry Class-G cementS2: Neat Class-G cement slurry,
w/c=0.44S3: Class-G cement slurry with 3% LS-A, w/c=0.44
S3A: Ca(OH)2B: AFtC: C3SD: C2S
D
DD
D
C
CC
C
C
BBB A
AA
AA
, degrees Fig. 4 XRD patterns of cement samples after 48 hours
of hydration at 4 C
10 20 30 40 50
D
D
D
D
C
C
C
C
C
BB
BA
A
AA
A
S2: Class-G cement, w/c=0.44S3: Class-G cement slurry with 3%
LS-A, w/c=0.44
A: Ca(OH)2B: AFtC: C3SD: C2S
2, degrees
S3
S2
Fig. 5 SEM images of hydration products of set cement with 3%
LS-A
(a)
(c)
CSH CSH
Ca(OH)2
Ca(OH)2AFt
CSH
(d)
(b)
SFI 3.0kv X10.000 1m WD5.9mm SEI 3.0kv X50,000 100nm WD5.9mm
SEI 3.0kv x50,000 100nm WD6.0mm SEI 3.0kv X50,000 100nm
WD5.9mm
Pet.Sci.(2011)8:328-334
-
332
of bar-shaped ettringite AFt crystals, which are the same as
those produced from the neat Class-G cement slurry.
4 Effect of LS-A on the properties of the cement slurry 4.1
Thickening time of the cement slurry
In deepwater surface-casing cementing operations, the
circulating temperature gradually declines due to the temperature
gradient and the convective heat transfer between the cold sea
water and the riser. The circulating temperature of the cement
slurry in deepwater wells is within the range of 10 to 20 C. The
thickening time of the cement slurry with 3% LS-A was measured at
10 C/7 MPa, and 15 C/10 MPa, respectively, as shown in Fig. 6.
Results show the accelerator LS-A can reduce the thickening time to
4-5 hours at low
Fig. 6 Effect of LS-A on thickening time of the Class-G cement
slurry
0 30 60 90 120 150 180 210 240 270 3000
30
60
90
120
150
180
210
240
270
Tem
pera
ture
,;
Pre
ssur
e,M
Pa
0
15
30
45
60
75
90
105
120
135
Time, min
(100 Bc: 307 min)
(30 Bc: 257 min)
: Temperature: Pressure : ConsistencyThickening time: 307
min
Con
sist
ency
, BC
(a) 10 C/7 MPa
0 30 60 90 120 150 180 210 240 2700
30
60
90
120
150
180
210
240
270
Time, min
(100 Bc: 248 min)
(30 Bc: 195 min)
0
15
30
45
60
75
90
105
120
135
: Temperature: Pressure: ConsistencyThickening time: 248 min
Tem
pera
ture
, ; P
ress
ure,
MP
a
Con
sist
ency
, Bc
(b) 15 C/10 MPa
temperatures, which is helpful to ensure safe cementing
operations and a short waiting-on-cement (WOC) time. The measured
thickening time of the neat cement slurry at 15 C/10 MPa, however,
was approximately 15 hours. Compared with the thickening time of
the cement slurry containing 3% CaCl2, which was cured at 15 C/10
MPa (Fig .7), the addition of LS-A had no unfavorab le effect on
the initial consistency of the cement slurry, but the addition of
CaCl2 led to a sharp increase in the consistency above 50 Bc at 12
min and a rapid decrease to 25 Bc at 21 min, this rapid hydration
is known as fl ash set. Moreover, for LS-A-containing slurry
system, the transition time that the consistency increased from 30
to 100 Bc was 53 min, but for the CaCl2-containing slurry system it
was about 98 min. This indicates that LS-A was helpful to shorten
the transition time from 30 to 100 Bc of the cement slurry.
4.2 Static gel strength of the cement slurryIn deepwater
surface-cementing cementing operations,
the casing setting depth is always near to 1,000 m below the mud
line, where the bottom hole static temperature is about 30 C
(OLeary et al, 2004; Tahmourpour and Quinton, 2009). To reveal the
effect of LS-A on the cement slurry to prevent upward fluid flow,
such as gas migration and fluid flows, through and along the cement
slurry, the static gel strength was measured at 30 C/14 MPa,
respectively, for the neat cement slurry, cement slurry with 3%
CaCl2, and the cement slurry with 3% LS-A. Fig. 8 indicates that
the time required for the cement slurry to change from having a
static gel strength of 0 Pa to having a static gel strength of 576
Pa was 251 min, and the transition time during which the static gel
strength increased from 48 to 240 Pa was 82 min. The addition of 3%
CaCl2 in the cement slurry shortened the time required for the
slurry to go from a static gel strength of 0 to 576 Pa, to 190 min,
but because of the fl ash set caused by CaCl2, the static gel
strength of the cement slurry rapidly reached 48 Pa at 14 min. This
is consistent with what is shown in Fig. 7, in which the cement
slurry with CaCl2 had a
high consistency of about 25 Bc after 12 min. The static gel
strength of CaCl2-containing cement system (in Fig. 8) was as high
as about 48 Pa after 14 min, and the transition time
Fig. 7 Effect of CaCl2 on thickening time of the Class-G cement
slurry at 15 C and 10 MPa
0 30 60 90 120 150 180 210 240 270 3000
30
60
90
120
150
180
210
240
270
0
15
30
45
60
75
90
105
120
135
(100 Bc: 287 min)
: Temperature: Pressure : ConsistencyThickening time: 287min
(30 Bc: 189 min)
Time, min
Tem
pera
ture
, ; P
ress
ure,
MP
a
Con
sist
ency
, Bc
Pet.Sci.(2011)8:328-334
-
333
of critical gel strength from 48 to 240 Pa was 116 min, 1.41
times longer than that of the neat cement slurry. However, the
addition of 3% LS-A shortened the time of gel strength from 0 to
576 Pa and transition time of critical gel strength from 48 to 240
Pa to be 122 min and 31 min, respectively.
Experimental results indicates that the LS-A signifi cantly
promotes the development of static gel strength of the Class-G
cement and shortened its transition time required to go from a gel
strength of 48 Pa to 240 Pa. With respect to the mechanism to
control gas or fluid migration during cementing, a short transition
time required for the cement slurry to go from a critical gel
strength of 48 Pa to 240 Pa is of helpful to prevent fl uid or gas
intrusion from the reservoir and prevent upward fl uid fl ow along
the cement slurry (Rogers et al, 2004). The above-mentioned
analysis demonstrates that the LS-A enhanced the cement slurry to
control gas or fl uid intrusion into the cement column.
cement at low temperatures. Low temperature leads to a greatly
increase in the 12- and 24-hours compressive strength of the
LS-A-containing cement system, which indicates that LS-A can
signifi cantly accelerate the early compressive strength
development of cement slurries at low temperatures.
0 20 40 60 80 100 120 140 160 180 200 220 240 2600
50
100
150
200
250
300
350
400
450
500
550
600
Neat3% CaCl23% LS-A
Class-G cement slurries with different acceleratorsw/c=0.44
95 min
64 min
130 min
14 min
180 min
Sta
tic g
el s
treng
th, P
a
Time, min
98 min
Fig. 8 Development of static gel strength as a function of time
for Class-G cement slurries
4.3 Compressive strength of the set cementFig. 9 presents the
compressive strength development
of the neat cement slurry, and slurry with 3% LS-A. Experimental
results indicate that the neat cement slurry exhibited no
compressive strength development when cured at low temperatures (4
and 10 C) for 12 hours, which is consistent with the XRD analysis
results. This means that Class-G cement particles will not hydrate
at 4 C for 12 hours. The cement slurry still had a very low
compressive strength (0.83 MPa) after being cured for 24 hours.
However, the set cement containing 3% LS-A developed a larger early
compressive strength at 12 hours (0.7 MPa) compared with the neat
cement slurry and was still able to further significantly increase
its compressive strength in the later stage. The 24-hour
compressive strength values of the LS-A-containing cement system
cured at 4, 10, and 20 C were respectively 11.36, 10.69, and 3.33
times larger than that of the neat cement slurry. The 48-hour
strength was respectively 2.06, 2.50 and 1.84 times that of the
neat Class-G cement slurry. The LS-A is proved to be capable of
significantly enhancing the early strength development of the
Class-G
0
5
10
15
20
25
30
24.83
8.24
13.52
20.58
7.88
3.82
17.10
5.14
8.87
0.832.84
0.25
8.90
0.52
2.300.70
48 hours24 hours12 hoursCuring time
Com
pres
sive
stre
ngth
, MP
a
Class-G set cement with different acceleratorsw/c=0.44
Neat, at 4 C 3% LS-A, at 4 C
Neat, at 10 C 3% LS-A, at 10 C
Neat, at 20 C3% LS-A, at 20 C
Fig. 9 Development of the compressive strength of set cement
5 Conclusions1) The set-accelerating admixture comprising of
lithium
chloride, aluminium hydroxide, and alkaline metal chlorides,
named as LS-A, could speed up the hydration of C3S and C2S at
low-temperatures, and thereby improve the low-temperature
properties of the Class-G cement by accelerating the breakdown of
the protective film, and shortening the hydration induction
period.
2) LS-A had no effect on the types of hydration products of the
Class-G cement. The products were still calcium silicate gel CSH,
Ca(OH)2 crystals, and a small number of AFt crystals.
3) LS-A significantly shortened the thickening time and enhanced
the early strength development of the Class-G cement slurry at low
temperatures. It exhibited excellent low-temperature acceleration
and early strength enhancement, and could effective reduce the
waiting-on-cement time to within 12 hours.
4) LS-A shortened the transition time of critical gel strength
from 48 to 240 Pa, so it is helpful to minimize the chance of
shallow water/gas migration. LS-A had superior performance over
existing accelerators, so it is a practical and reliable new
accelerator for deepwater cementing.
AcknowledgementsThe fi nancial support is provided by the Ph.D.
Programs
Foundation of Ministry of Education of China (Grant No.
20100133120004), National Major Science and Technology Project of
China (Grant No. 2009ZX05060) and National High Technology Research
and Development Program of China (863 program, Grant No.
2006AA09Z340). The authors would also like to thank the anonymous
reviewers for their valuable suggestion and comments.
Pet.Sci.(2011)8:328-334
-
334
ReferencesBro thers L E and Palmer A. Cementing in deepwater
offshore wells. US
Patent: 6835243 B2, December 28, 2004 Dou ble D D and Hellawell
A. The hydration of Portland cement. Nature.
1976. 261(10): 486-488Gri ffi th J E. Method for cementing in a
formation subject to water infl ux.
US Patent: 5484019A, 1996-1-16Kje llsen K O and Justnes H.
Revisiting the microstructure of hydrated
tricalcium silicate: A comparison to Portland cement. Cement and
Concrete Composites. 2004. (26): 947-956
Kje llsen K O and Lagerblad B. Microstructure of tricalcium
silicate and Portland cement systems at middle periods of
hydration-development of Hadley grains. Cement and Concrete
Research. 2007. (37): 13-20
Lu P, Zhai J P, Nie R, et al. Investigation on the early stage
hydration of Portland cement using environmental scanning electron
microscopy. Journal of the Chinese Ceramic Society. 2004. 32(4):
530-536 (in Chinese)
Ma berry J, Garrison G, Harris S, et al. Process of well
cementing in cold environment. US Patent: 6955220B2, October 18,
2005
Nel son E B. Well Cementing. Netherlands: Schlumberger
Educational Services. 1990. 43-52
OL eary J, Flores J C, Rubinstein P and Garrison G. Cementing
deepwater, low-temperature Gulf of Mexico formations prone to
shallow flows. Paper SPE 87161 presented at SPE Drilling Conference
held in Dallas, Texas, March 2-4, 2004
Pet tingill H S and Weimei P. World-wide deepwater exploration
and production: Past, present and future. Paper SPE 14024 presented
at the Offshore Technology Conference, 6-9 May 2002, Houston,
Texas
Rae P and Lullo G D. Lightweight cement formulations for deep
water cementing: Fact and fiction. Paper SPE 91002 presented at
SPE
Annual Technical Conference and Exhibition held in Houston,
Texas, September 26-29, 2004
Rav i K, Biezen E N, Lightford S C, et al. Deepwater cementing
challenges. Paper SPE 56534 presented at SPE Annual Technical
Conference and Exhibition held in Houston, Texas, October 3-6,
1999
Rav i K M, Gastebled O and Bosma M G R. Method for selection of
cementing composition. US Patent: 6922637B2, July 26, 2005
Red dy B R and Fitzgerald R M. Cementing casing strings in
deepwater offshore wells. US Patent: 6454004 B2, September 24,
2002
Rog ers M J, Dillenbeck R L and Eid R N. Transition time of
cement slurries, definitions and misconceptions, related to annular
fluid migration. Paper SPE90829 presented at SPE Annual Technical
Conference and Exhibition held in Houston, Texas, September 26-29,
2004
Tah mourpour F and Quinton C. Saving expensive offshore
deepwater rig time by modeling accurate subsea/subsea-floor
temperature modeling for cementing operations. Paper SPE123738
presented at SPE Annual Technical Conference and Exhibition held in
New Orleans, Louisiana, October 4-7, 2009
Wan g C W, Wang R H, Bu Y H, et al. Research development of
deepwater cementing technique. Journal of China University of
Petroleum (Edition of Natural Science). 2008. 32(1): 77-81 (in
Chinese)
Wan g C W, Wang R H, Bu Y H, et al. Properties and hydration
mechanism of deepwater cementing system. Acta Petrolei Sinica.
2009. 30(2): 280-284 (in Chinese)
Yu C J, Mo X Y, Deng M, et al. Behavior characteristic of
lithium ion in alkali-aggregate reaction. Journal of Southeast
University (Natural Science Edition). 2009. 39(1): 127-130 (in
Chinese)
(Edited by Sun Yanhua)
Pet.Sci.(2011)8:328-334
Abstract:Key words:1 Introduction2 Experimental2.1 Materials2.2
Methods
3 Acceleration mechanism of LS-A3.1 Effect of LS-A on
low-temperature hydration of oil well cement3.2 Effects of LS-A on
low temperature hydration products3.3 Effects of LS-A on the
microstructure of hydration products
4 Effect of LS-A on the properties of the cement slurry4.1
Thickening time of the cement slurry4.2 Static gel strength of the
cement slurry4.3 Compressive strength of the set cement
5 ConclusionsAcknowledgementsReferences