Note: Descriptions are shown in the official language in which they were submitted.
10LOW DENSITY CEMENT SLURRY AND ITS USE
15SUMMARY OF THE INVENTION
In oil well type completions of casing strings,
weak formations are encountered which necessitate the use
of lightweight cement. The lightweight cement of this
invention comprises hydraulic cement, hollow glass micro-
20 spheres and sufficîent water to form a pumpable slurrywith an API free water content of no more than about 2
volume percent. These microspheres have true particle
densities of about 0.2 to about 0.S gm/cm~ as determined
by ANSI/ASTM D 2840-69, hydrostatic collapse strengths of
25 at least about 500 psi (3447 kPa) as determined by
AN~I/ASTM D 3102-72 and average particle diameters of less
than about 500 microns and can be used at about 8 to about
50 weight percent based on the weight of the hydraulic
cement to satisfactorily produce slurries having densities
30 of less than about 12 ppg (1.44 kg/litre). The light-
weight cements of this invention have lower densities and
attain higher strengths than the previously used cementing
compositions with water added to reduce the density of the
compositions and materials such as bentonite, diatomaceous
35 earth, or sodium metasilicate added to keep the composi-
tions from separating.
,.,
~ .
--2--
~RIEF DESCRIPTION OF THE DR~WINGS
Figure 1 is a comparison of the strength of the
cement of this invention to the strength of previously
used cements.
Figure 2 shows the strength of hollow glass
microspheres used in th:is invention.
Figure 3 shows the increase in slurry density of
a slurry of this invention as hydrostatic pressure is
increased.
Figure 4 shows the water required to form the
slurries of this invention.
Figure 5 shows the slurry consistency of the low
density cement of this invention.
Figure 6 shows the slurry densities attained by
15 adding hollow glass spheres.
Figure 7 shows API thickening times of slurries
of this invention.
Figure 8 shows API compressive strengths of
slurries of this invention.
DETAILED DESCRIPTION
In oil well type cementing, it has now been
found that a low density cement which comprises hydraulic
cement, hollow glass microspheres and water is superior to
previously described lightweight cements. Low density or
25 lightweight cements are used in the completion of wells
which extend through weak subterranean formations to
reduce the hydrostatic pressure exerted by a column of the
cement on the weak formations. Examples of such forma-
tions are the unconsolidated Late Tertiary formations
30 encountered in the Gulf Coast Region of the United States,
shallow coal seams encountered in Wyoming, Muskeg forma-
tions encountered in Canada, and fractured formations
encountered worldwide. These formations are encountered
when drilling wells for the recovery of subterranean
35 resources such as oil, gas, minerals and water and the
lightweight cement of this invention is useful in complet-
ing these wells. These completions are referred to herein
as oil well type completions and include but are not lim-
ited to completions where the cement slurry is pumped
--3~
downwardly through the casing in a well and upwardly into
the annulus between the casing and the wall of the well,
the cement slurry is placed in the annulus between the
casing and the wall of a well by grouting techniques and a
5 plug of the cement slurry is placed in the well for aban-
donment or for establishing a whipstock.
It is described in Part 3 of a series of arti-
cles on Basic Cementing, Oil and Gas Journal, Volume 75,
No. 11, March 14, 1977, that "Basically, lightweight slur-
10 ries are made by adding more water to lighten the mixtureand then adding materials which keep the solids from sepa-
rating." Bentonite, diatomaceous earth and sodium
metasilicate are described as materials which can be added
to keep the solids from separating when the water is added
15 for reducing the slurry density. It is also described
that slurry densities as low as 10.8 ppg (1.29 kg/litre)
can be achieved by adding water. This method of producing
lightweight cement slurries has the drawback that the
addition of more water increases the cure time and reduces
20 the strength of the resulting lightweight cement to the
extent that they cannot be mixed to densities of less than
about 10.8 ppg (1.29 kg/litre).
The low density cements of this invention are
made by adding hollow glass microspheres and sufficient
25 water to hydraulic cement to form a pumpable slurry. The
additional water required because of the use of the micro-
spheres is substantially less than that required when
water is added to reduce the slurry density; therefore,
the set cement of this invention has a substantially
30 higher strength than previously used low density cements
and cure time is reduced. Because of the additional
strength, the lightweight cement of this invention can be
formulated at densi.ties much lower than the slurry density
of 10.8 ppg (1.29 kg/litre) as described in the above
35 referenced publication. This is illustrated in Figure 1
where the compressive strengths of lightweight cements
described in the above referenced publication are compared
with the compressive strengths of the lightweight cements
of this invention.
--4--
The comparisons shown in Figure 1 were made with
American Petroleunl Institute (API) Class A cements at the
API well simulation test schedule for a 1,000 foot (305
meter) well as described in "API Recommended Practice for
5 Testing Oil Well Cements and Cement Additives," American
Petroleum Institute, Washington, D.C., ~PI RP-lOB, 20th
Edition, April 1977. These comparisons can be made with
respect to the above referenced Oil and Gas Journal publl-
cation where it is described that most operators wait for
10 cement to reach a minimum compressive strength of 500 psi
(3447 kPa) before reswming operation. It is seen in Fig-
ure 1 that a 9 ppg (1.08 kg/litre) cement mixture of this
invention will attain a compress~ve strength of about SOO
psi (3447 kPa) in 24 hours as compared to a 13 ppg (1.56
15 kg/litre) cement mixtllre using water to reduce the density
of the cement mixture and bentonite, diatomaceous earth or
sodium metasilicate to keep the solids from separating.
It is also seen that an 11 ppg cement mixture of this
invention will attain a compressive strength of about 500
20 psi (3447 kPa) in 12 hours. This illustrates that the use
of the lightweight cement of this invention can consider-
ably reduce the waiting time while the cement is curing to
the minimum strength, waiting on cement time (WOC).
The lightweight cement of this invention is also
25 superior to the lightweight cement described in Biederman,
U.S. 3,699,701, Maxson, U.S. 3,722,591, Gebhardt,
U.S. 3,782,985, and Messenger U.S. 3,804,058. It is des-
cribed in column 2 at lines 35 through 38 of Biederman
that lightweight oil well cements may be formed by the
30 addition of float ash, as aggregate, to existing oil well
cements. Float ash is described in column 1 at lines 13
through 15 of Biederman as the portion of fly ash that
floats on water and has a specific gravity around 0.7.
The use of fly ash floaters is also described in column 2
35 at lines 7 through 18 of Gebhardt. It is described in
column 2 at lines 17 through 22 of Maxson that a borehole
can be lined with an insula-ting liner formed in situ by
hardening in place a hardenable, flowable composition con-
sisting essentially of a hardenable, flowable, adhesive
-5-
cement and a divided, solid, closed-cell material. A sui-
table aflhesive cement is described in column 3 at line 18
of Maxson as hydraulic cement. A suitable divided, solid,
closed~cel] material is described in co:Lumn 3, at lines 59
5 through 63 of Maxson as fly ash floaters. It is described
in column 5, at lines 4 through 8 of Messenger that pump-
able cement slurries can be produced by mixing portland
cement, anhydrous sodium metasilicate, water, and hollow
sealed spheres made of ceramic or glass. Ceramic and
10 glass spheres are compared in column 5 at lines 30 through
37 of Messenger where it is described that ceramic spheres
are preferred in wells having hydrostatic pressures upward
to about 2500 psi (17,237 kPa). The water required for
producing the pumpable slurry of Messenger is described in
15 column 5 at lines 38 through 55 and in column 6 at lines 1
and 2, where it is described that additional water is
added for the glass spheres and for the sodium metasili-
cate. Advantages achieved by adding this extra water are
described in column 2 at lines 11 through 16 of Messenger
20 where it is described that the extra water increases the
space between the suspended hollow sealed glass spheres
and thus reduces the breakage of the spheres.
The additional water required by Messenger would
have the same detrimental effects as described with
25 respect to the lightweight slurries produced by adding
water to reduce the density of the slurries and bentonite,
diatomaceous earth, or sodium metasilicate to keep the
solids from separating. Sample numbers ~a, 8, 9, 9a, 9-b
and 9c in Table II of Messenger are examples of Messen-
30 ger's use of additional water along with sodium metasili-
cate to prevent water separation. Messengers addition of
more water increases the cure time and reduces the
strength of the resulting lightweight cement. Therefore,
the lightweight cement of Messenger has longer cure times
35 and is inferior in strength to the cement slurries of the
present invention which are free from effective amounts of
additives such as bentonite, diatomaceous earth and sodium
metasilicate. An effective amount of one of these addi-
tives is an amount which would prevent the separation of
?''~Q
-6
water and particles from the slurry.
The API free water content of sl~rries of port-
land cement, IG 101 glass spheres marketed by Emerson &
Cuming, Inc., and the amount of water described by Messen-
5 ger in column 5 at lines 35 through 38 and column 6 atlines 1 and 2, and shown in Table 2, is too high for oil
well type completions.
The API free water content of a slurry is main-
tained at no more than 2 volume percent and preferably
10 less than about 1 volume percent to minimize water separa-
tion after placement of the slurry. Water separation in a
column of cement can form pockets of free water within the
cement column or reduce the height of the column of
slurry. Pockets of free water within a cement column can
15 cause corrosion of adjacent casing. Higher percentages of
water separation can form channels through which fluid can
migrate past the cement column.
The API free water content of a slurry prepared
by mixing API Class H cement with 20.68 weight percent IG
20 101 hollow glass spheres and 126 weight percent water is
about 18.4 volume percent. In addition to the water sepa-
ration, solids segregated during this API test. The lower
portion of the 250 milliliter graduated cylinder in which
this test was conducted contained 84 milliliters of
25 solids, the middle portion contained about 46 milliliters
of water and the top portion contained about
120 milliliters of solids floating on the water. This
slurry has the same components as sample No. l in Table 2
of Messenger. The free water content of another cement
30 slurry prepared by mixing API Class H cement with 10.34
weight percent IG 101 hollow glass spheres and 84 weight
percent water is about 16 volume percent. This slurry
also suffered about the same severe particle separation as
described with respect to the slurry mixed with 20.68 per-
35 cent spheres. This slurry contains the amount of waterspecified in column 5 at lines 53 through 55 for the
cement and glass spheres. It is described in Messenger
that a slurry should contain 5 gallons of water per 94
pounds of cement and 4.4 gallons of water per each 10
weight percent glass spheres. This is equivalent to for-
mulating the slurry with about 44 weight percent water for
the cement and about 3.9 weight percent water for each
weight percent glass spheres. A slurry mixed with this
- 5 amount of water would contain about 84 weight percent
water when mixed with about 10.34 weight percent glass
spheres and about 124 weight percent water when mixed with
about 20.68 weight percent glass spheres. These weight
percents are based on the weight of the cement.
The ceramic spheres described in Messenger are
marketed by Emerson and Cuming, Inc under the trade desig-
nation FA-A Eccospheres. These ceramic spheres are known
- in the industry as fly ash floaters or float ash.
It has been found that hollow ceramic spheres
15 have properties which render them undesirable as additives
in lightweight cements for oil well completions. A 12.7
ppg (1.52 kg/litre) slurry mixed with API Class A cement,
about 20 weight percent hollow ceramic spheres and suffi-
cient water to form a slurry with an API normal water con-
20 tent lost about 10% of its volume when subjected to pres-
sures of less than about 5000 psi (34,474 kPa) and a 10
ppg (1.2 kg/litre) slurry mixed with API Class A cement,
about 67 weight percent hollow ceramic spheres and suffi-
cient water to form a slurry with an API normal water con-
25 tent lost about 10% of its slurry volume when subjected topressures of less than about 2000 psi. The loss of volume
by the 10 ppg (1.2 kg/litre) slurry results in a solid
plug that is not pumpable. Slurries of this invention do
not lose their pumpability after loss of 10% of their
30 slurry volume under hydrostatic pressure, additionally the
loss of slurry volu~le of the slurries of this invention is
not proportional to the amount of hollow spheres mixed
with the slurry.
It has also been found that lightweight cements
35 mixed with these ceramic spheres suffer substantial
increases in slurry density and reductions in slurry
volume when cured under a hydrostatic pressure of about
500 psi (3447 kPa). A 9.7 ppg (1.16 kg/litre) slurry of
API Class A cement, ceramic spheres, and sufficient water
-8--
to form a slurry with an API normal water content did not
shrink or increase in densi. Ly when cured at atmospheric
temperature and pressure. However, a sample of the same
slurry cured at atmospheri.c temperature and under a hydro-
5 static pressure of 500 psi (3~47 kPa) lost about 10% ofits volume and increased in density to about 10 ppg (1.2
kg/litre).
The effect of hydrostatic pressure on the volume
of a lightweight cement slurry mixed with hollow ceramic
10 spheres detracts from the use of these slurries in oil
well type completions. These detrimental effects are par-
ticularly serious when lightweight cement slurries are
mixed with hollow ceramic spheres to produce slurries
having densities of less than about 10 ppg (1.2 kg/litre).
15 The lightweight cement of this invention does not suffer
the substantial detrimental effects as shown in this
application for lightweight cement slurries produced with
hollow ceramic spheres. It will be described with respect
to Figures 2 and 3 of this disclosure that the effect of
20 hydrostatic pressure on the slurries of this invention is
proportional to the density of the hollow glass spheres
selected.
The hollow glass microspheres used to produce
the lightweight cement of this invention have average par-
25 ticle diameters of less than about 500 microns and can bemanufactured by the procedures described in Beck, et al,
U.S. 3,365,315 and Veatch, et al, U.S. 3,030,215. In
Beck, et al, it is described that hollow glass micro-
spheres are made by passi.ng particles of glass containing
30 a gas forming material through a current of heated air or
a flame. The gas forming materials can be incorporated
within the glass particles by the simple step of allowing
the particles, either at room temperature or at higher
temperatures below melting, to absorb or adsorb the fol-
35 lowing materials from the atmosphere surrounding the par-
ticles: H2O, CO2, SO2, F2, etc. It is described in
Veatch, et al, that hollow glass microspheres are made by
subjecting a particulated mixture of siliceous material
such as sodium silicate, a water-desensitizing agent such
-9-
as boric acid and a blowing agent such as urea to an ele-
vated temperature for a time necessary to fwse the parti-
cl.es and cause expansion of the particles to hollow glass
spheres, High strength glass such as borosilicate glass
5 can be used to produce hollow microspheres having
hydrostatic collapse strengths of greater than about 5,000
psi (34474 kPa~ as determined by the American Society for
Testing and Materials procedure described in ANSI/ASTM D
3102-72.
It is illustrated in Figure 2 that hollow micro-
spheres with average true particle densities of about 0.2
gm/cm3 as determined by the American Soci.ety for Testing
and Materials procedure described in ANSI/ASTM ~ 2840-69
generally have ANSI/ASTM hydrostatic collapse strengths of
15 less than about 500 psi (3447 kPa) while hollow glass
microspheres with ANSI/ASTM average true particle densi-
ties of about 0.5 gm/cm3 have ANSI/ASTM hydrostatic col-
lapse strengths of greater than about 5,000 psi (34474
kPa).
The hydrostatic collapse strength measurements
shown in Figure 2 were made in accordance with the proce-
dure described in ANSI/ASTM D 3102-72. The pressure
required to collapse about 10 volume percent of the hollow
glass microspheres is reported as the hydrostatic collapse
25 strength of the microspheres. This is thought to simulate
the hydrostatic pressure under which cement slurries con-
taining these microspheres will be subjected during oil
well type cementing operations and also simulates the
isostatic pressure under which these microspheres will be
30 subjected during these operations. These microspheres are
generally manufactured to have average particle diameters
of about 10 to about 300 microns.
Many of the uses for the lightweight cements of
this invention will be in the completion of wells having
35 depths of about 1,000 feet (305 meters) to about 6,000
feet (1,830 meters). It is described in API RP-lOB tha-t
wells, for API simulated test conditions, having depths of
about 1,000 feet (305 meters) will have bottom hole pres-
sures of about 1020 psi (7,000 kPa) and that wells having
--1 0
depths of about 6,000 feet (1,830 meters) will have bottom
hole pressures of about 3870 psi (26,700 kPa). Commer-
cially available hollow glass microspheres with ANSI/ASTM
true particle densities of about 0.3 gm/cm3 and ANSI/ASTM
5 hydrostatic collapse strengths of about 1,000 psi(6895
kPa) are satisfactory for completing API simulated wells
to depths of about 1,000 feet (305 meters). Commercially
available hollow glass microspheres with ANSI/ASTM true
particle densities of about 0.4 gm/cm3 and ANSI/ASTM
10 hydrostatic collapse strengths o~ about 4,000 psi (27579
kPa) are satisfactory for completing API simulated wells
to depths of about 6,000 feet (1,830 meters). In general,
higher density and thus higher strength hollow glass
microspheres are desirable for completing deeper wells.
15 Glass spherès having ANSI/ASTM hydrostatic collapse
strengths of greater than about 4000 psi (27,579 kPa) may
be more economical for completing wells to depths of
10,000 feet (3,050 meters) and deeper.
It has been observed that cement slurries of
20 this invention increase in density as pressure on the
slurry is increased. This is illustrated in Figur~ 3
where it is shown that slurry density can be compensated
for by mixing the slurry with an amount of hollow glass
microspheres which will provide the appropriate slurry
25 density under the hydrostatic conditions which the slurry
will be subjected. It is shown that a slurry to be sub-
jected to a hydrostatic pressure of about 2,000 psi (13790
kPa) should initially contain a higher concentration of
these microspheres than a slurry to be subjected to a
30 hydrostatic pressure of about 1,000 psi (6895 kPa) and
that after being subjected to these maximum hydrostatic
pressures, both slurries would contain about the same con-
centration of these microspheres as evidenced by their
slurry densities. It is noted from Figure 3 that a higher
35 percentage of the microspheres is lost as the hydrostatic
pressure is increased from 1,500 to 2,000 psi (10342 to
13790 kPa) than is lost as the hydrostatic pressure is
increased from 500 to 1,000 psi (3447 to 6895 kPa) The
microspheres used for the tests shown in Figure 3 would be
-11 -
satisfactory for completing a well with a bottom hole
hydrostatic pressure of about 1,000 psi (6895 kPa); how-
ever, microspheres with higher collapse strengths may be
more economical for use at greater depths.
The low density cement of the present invention
is mixed with sufficient water to form a pumpable slurry
with a free water content of no more than about 2 volume
percent and preferably less than about 1 volume percent as
determined by the procedure described in Section 4 "Deter-
10 mination of Water Content of Slurry," API RP-lOB. The
pumpable slurry of this invention preferably has greater
than the minimum water content and most preferably has
about the normal water content, both as described in this
section of API RP-lOB. A slurry having a minimum water
15 content is described in API RP-lOB as having a consistency
of about 30 Bearden units of slurry consistency (Bc) while
a slurry having a normal water content is described as
having a consistency of about 11 Bc. A slurry with less
than an API minimum water content is difficult to pump and
20 a slurry with an API normal water content is considered as
having an optimum consistency for pumping and a satisfac-
tory free water content.
It is illustrated in Figure 4 that a slurry of
this invention formulated with API Class A portland cement
25 and about 8 to about 50 weight percent hollow borosilicate
glass microspheres based on the weight of the cement can
be mixed with sufficient water to provide a slurry with an
API normal water content. The hollow glass microspheres
illustrated in these tests are B37/2000 microspheres mark-
30 eted by the Minnesota Mining and Manufacturing Company andhave average particle diameters of about 20 to about 130
microns. Other microspheres may have other water requi.re-
ments. This slurry is mixed with sufficient water to form
a pumpable slurry with the portland cement and an addi-
35 tional amount of water because of the microspheres equalto about 1.2 weight percent extra water for each weight
percent of the hollow microspheres when the slurry is
mixed with about 8 weight percent of the microspheres
based on the weight of the cement and about 2.2 weight
~z~
-12-
percent extra water for each weight percent of these
microspheres when the slurry is mixed with about 50 weight
percent of these microspheres by weight of the cement. At
about 10 weight percent of these microspheres, the API
5 free water content of a slurry mixed with about 1.3 weight
percent extra water for each weight percent hollow micro-
sphere is about 2.5 milliliters or about 1 volume percent.
At about 30 weight percent of these microspheres the API
free water content of a slurry mixed with about 1.8 weight
10 percent extra water for each weight percent hollow micro-
sphere is about 1 milliliter or about 0.4 volume percent.
The weight percent water for each weight percent
of microspheres as shown in Figure 4 is in addition to the
water required to give a pumpable slurry with the cement.
15 This is illustrated by an API Class A portland cement
mixed with about 10 weight percent hollow glass micro-
spheres having ANSI/ASTM average true particle densities
of about 0.37 gm/cm3 and about 59 weight percent water
based on the weight of the cement to give a slurry having
20 a density of about 12 ppg (1.4 kg/litre) and an API normal
water content. About 46 weight percent water based on the
weight of the cement is required to give a pumpable slurry
with the cement and about 13 weight percent water based on
the weight of the cement is required to wet the surface of
25 the glass spheres and to give a pumpable slurry with the
mixture of cement and glass spheres. Without the addi-
tional water for each weight percent of microspheres, the
slurry of cement, glass spheres and water may not be pump-
able.
The water content of the hydraulic cement slur-
ries for use in producing the mixture of this invention is
given in Section 4 of API RP-lOB where normal and minimum
water contents are specified. The hydraulic cement can
contdin any conventional additives needed to meet well
35 conditions and should contain normal to minimum water con-
tents as required when such additives are mixed with the
hydraulic cement. Additives which may be desired are
accelerators, loss circulation materials and dispersants.
The following Table 1 also appears at Section 4 of API
-13-
RP-lOB and provides the water required for mixing neat API
Classes of cement with water to produce slurries with nor-
mal water contents.
TABLE I
CEMENT SLIJRRY COMPOSITION
1 2 - - 3
Water Percent Water
API Class by Weight ofGallons Litres
Cement Cement per _4 lb Sack per 42.6 kg Sack
A & B 46 5.19 19.6
15 C 5~ 6.32 23.9
D, E. F, & H 38 4.29 16.2
G 44 4.97 18.8
J ,L ,L ,L
20 *As recommended by the manufacturer.
It is illustrated in Figure 5 that about 1.3
weight percent extra water for each weight percent
B37/2000 microsphere based on the weight of the cement
will provide a slurry having an API normal water content
25 at about 10 volume percent hollow microspheres based upon
the weight of cement to a slurry having an API minimum
water content at about 40 weight percent hollow glass
microspheres.
It is shown in Figure 6 that an API Class A
30 cement mixed with glass microspheres having ANSI/ASTM
average true particle densities within the range of about
0.2 to about 0.5 gm/cm3 and an API normal water content
can be formulated to produce slurries having densities of
less than about 12 ppg (1.4 kg/litre). Generally micro-
35 spheres with ANSI/ASTM true particle densities within therange of about 0.3 to about 0.4 gm/cm3 will be used to
produce slurries having densities of about 9 to about 12
ppg (about 1.08 to about 1.4 kg/litre).
~ s;,~
The lightweight cement of this invention can be
formulated with any hydraulic cement normally used in oil
well type cementing and the hollow glass microspheres can
be used in combination with other additives. Portland
5 cements are the basic hydraulic cements now being used for
oil well type cementing and are often mixed with accelera-
tors, retarders, dispersants and loss circulation agents
to meet specific well conditions.
The use of calcium chloride as an accelerator in
10 the low density cement of this invention is illustrated in
Figures 7 and 8 where it is seen that the thickening times
of the lightweight cements of this invention generally
increase as they are mixed with higher concentrations of
hollow glass microspheres and that their compressive
15 strengths generally decrease at higher concentrations of
microspheres. It is also seen that calcium chloride has a
greater accelerating effect on the lightweight cements of
this invention mixed with lower concentrations of hollow
glass microspheres. The calcium chloride is diluted by
20 the additional water for each weight percent glass micro-
sphere.
Tests have also been conducted on API Class A
cement mixed with about 25.5 weight percent hollow glass
microspheres having ANSI/ASTM true particle densities of
25 about 0.37 gm/cm3 and with dispersant and gypsum hemi-
hydrate. The percent microspheres is based on the weight
of the cement. Dispersants are known for reducing the
water required in the mixing of cements and gypsum hemi-
hydrate is known for producing a cement slurry that will
30 attain a high gel strength when movement of the slurry is
reduced or terminated. Cement slurries which gel when
movement is reduced are useful for plugging fractures or
filling voids which extend from a wellbore and for reduc-
ing the hydrostatic pressure applied by the slurry to weak
35 subterranean formations after placement of the cement
slurry has been completed. The addition of 0.75 weight
percent dispersant based on the of weight of the cement
reduced the extra water required because of the addition
of the hollow glass microspheres from 1.65 to about l
. . ,
-15-
weight percent water for each percent of these micro-
spheres based on the weight of the cement. The adclition
of 7.5 weight percent gypsum hemihydrate provided a slurry
with a gel strength of about 1,0()0 pounds per 100 square
5 feet (50 kg/cm2) after movement of the slurry had been
suspended for about 12 minutes as compared to a gel
strength of about 50 pounds per 100 square feet (2.5
kg/cm2) for a slurry containing no gypsum hemihydrate.
The low density cement of this invention is also
10 useful in abnormally hot wells because it i.s characterized
on curing by having quite high compressive strength at
elevated temperatures (above 110C). In the unusual case
in which the API cement (except Class J) is to be cured
under or later subjected to high temperature conditions
15 (above around 110C), the initial compressive strength is
not maintained but may decrease rapidly, of the order of
30% or more. The lightweight cement of this invention is
also useful in steam or hot water injection wells and pro-
ducing wells from thermal sources and the like, as well as
20 in wells penetrating permafrost, where there is a definite
need to obtain a satisfactory insulating lining between
the fluid and the formations surrounding the well.
It has been observed that the lightweight cement
of this invention has a high temperature strength at temp-
25 eratures above 230F (110C) which is at least the sameorder of magnitude as (and sometime exceeds) that of the
material when cured at a temperature in the order of 90 to
120F (32 to 49C). This is illustrated with respect to
the following example of a slurry of the present inven-
30 tion. An API Class A cement slurry in the absence offinely divided hollow spheres would contain approximately
46 weight percent of water. If 25.5 weight percent of
B37/2000 spheres are added (ANSI/ASTM average average true
particle density of 0.37 gm/cm3), an additional 34 weight
35 percent of water, about 1.3 weight percent water for each
weight percent hollow spheres, for a total of 80 weight
percent water can be incorporated in this particular case.
The percentages are based on the weight of the cement.
The density of the resulting slurry is of the order of 9.5
-16-
ppg (1.14 kg/litre). When this slurry is cured at about
300F (149C) it has a one day strength of about 1,000 psl
(6895 kPa). A sample of this lightweight cement exposed
to a temperature of 300F (149C) for seven (7) days has a
5 compressive strength of about 970 psi (6688 kPa) which is
an inconsequential reduction. For all practical purposes,
one can say the two compressive strengths are identical.
Ordinarily, an API Class A oil well cement slurry with no
low density additive upon curing would have a l day
10 strength of the order of 3,000 psi (20684 kPa) at 300F
(149C) temperature, but exposed to this temperature of
300F (149C) for seven (7) days would regress typically
approximately 30%. A lightweight cement produced by
adding water to reduce the density of the slurry to about
15 13 ppg (1.56 kg/litre) and bentonite, diatomaceous earth
or sodium metasilicate to keep the solids from separating
would lose substantially all of its compressive strength
after being cured at 300F (149C) for one week.
Examples of the hollow glass microspheres which
20 have been found to be useful in formulating the light-
weight cement of this invention are shown in Table II.
! - 17-
TABLE II
Free-Flowing Hollow Glass Sphere Properties-Typical
Type lG101 lGD10l B23/500 B37/2000 B38/~lO00
5 Made by* E-C E-C 3M 3M 3M
Glass~L SB SB SLB SLB SLB
ANSI/ASTM average
true particle
density, (gm/cm3) 0.31 0.3 0.23 0.37 0.38
Average particle
diameter,
(microns) 40-175+ 40-150 20-130 20-130 20-130
ANSI/ASTM
hydrostatic collapse
strength at 10 vol.
percent collapse,
(psi) unknown unknown500 2000 4000
(kPa) unknown unknown3447 13790 27579
Hydrostatic collapse
strength - volume
25 percent survivors
at 1500 psi
(10340 kPa) 47 76.6unknown unknown unknown
Softening temperature
(C) 480 480 715 715 715
* E-C means Emerson & Cuming, Inc., Canton, Massachusetts
3M means 3M Manufacturing Co., St. Paul, Minnesota
~ - SB is sodium borosilicate glass
SLB is soda lime borosilicate glass
(Note: in the claims both sodium borosilicate
glass and soda lime borosilicate glass are referred to
.
,'
f~
-18-
generically as "sodium borosilicate glass" or "borosili-
cate glass." Thermal conductivity is of the order of 8 to
11 (K cal)(cm)(hr)(sq m)(C).)
The Emerson and Cuming microspheres are thowght
5 to be manufactured by the procedure described in Veatch,
et al, U.S. 3,030,215 and the 3M Manufacturing Company
microspheres are thought to be manufactured by the proce-
dure described in Beck, et al, U.S. 3,365,315. The com-
mercially available hollow glass microspheres manufactured
10 by the procedure described ln Beck, et al, have strengths
that vary with ANSI/ASTM true particle densities as shown
in Figure 2 and are useful under the broad range of hydro-
static pressure conditions expected to be encountered in
the use of a lightweight cement of this invention. Micro-
15 spheres manufactured by procedures other than the proce-
dures described in Beck, et al~ U.S. 3,365,315 may not be
satisfactory for use under conditions where the light-
weight cement of this invention is subjected to hydro-
static pressures of greater than about 1500 psi
20 (10340 kPa).
One note about use: since these tiny spheres
are made of glass~ it is apparent that one must be careful
in mixing the spheres and cement together. We do not find
it necessary to use metasilicate; however, we mix by mov-
25 ing these materials into a storage tank by use of some-
thing resembling a large vacuum cleaner or diaphragm pump,
and it is desirable that the people using these wear res-
pirators and goggles. There should be at least an ordi-
nary vacuum bag filter in the exit line from the vacuum
30 system. At the bulk station, where these mixtures are to
be mixed together, the spheres and cement are dry mixed in
the presence of sufficient air to cause homogeneous blend-
ing of the two materials. They are then moved into a
truck and sent to the well. At the well an ordinary Hal-
35 liburton jet mixer or equivalent can be employed to mixthese dry ingredients with water to the required density.
With the use of pneumatic bulk handling vessels, operators
may not be exposed to these tiny glass spheres and may not
need to exercise these safety precautions.
-19-
The use of a densltometer or a pressurized flwid
density balance as described in Appendix B to API RP-lOB
are about the minimum equipment currently necessary to
monitor the density. A manual centrifuge can also be used
5 in making the de-termination of water content. With a
manual centrifuge there is no need to depend upon a source
of electric power or the like. Shortly before the mixing
is going to commence at the field location, we prepare
small test samples containing the amount of solid ingre-
10 dients planned for a particular job, each having a calib-
rated amount of water amounting, for example, to 70%, 80%,
90%, and 100% (based on cement weight). These separate
samples are centrifuged which causes the material to sepa-
rate out into a cement portion, a water portion, and a
15 portion containing finely divided glass spheres. One pre-
pares a graph in which the true or calibrating percentage
of water is plotted along one axis and that determined
from the centrifuge volume is plotted along the other
axis. A smooth curve is drawn through the points. This
20 calibrates the centrifuge.
Then when the actual mixing of the cement, glass
spheres and water takes place, we take samples every two
minutes and centrifuge these in the calibrated centrifuge.
The volume of water found in the sample is read off
25 against the calibration curve to determine the actual per-
cent concentration of water in that particular sample.
The rapidity of the operation can be judged by the fact
that a typical figure for slurry mixing rate is about
barrels per minute t636 litre/min).
To give a specific example of such an operation,
2,000 lbs (907 kg) of IGD-101 hollow glass microspheres
were blended with 85 sacks, 8,000 lbs (3630 kg), of Okla-
homa API Class A cement. The cement and the tiny glass
hollow spheres were blended in two batches of equal
35 volume. In each batch, half the cement was vacuum
injected into the blender, then the spheres, and then the
remainder of this cement. The batches were blown back and
forth from tank to bulk truck three times to mix the hol-
low spheres with the cement. The blending operation took
-20-
1-1/2 hours, including unpacking the glass spheres. Four
cement company personnel were eMployed in this operation.
Dust was not excessive but goggles and dust masks were
worn by all personnel in the blending area. In this oper-
5 ation, the glass spheres were dumped into a cone very muchlike the Halliburton hopper for jet mixing, then were
sucked vertically out of the cone with a 6 in. (15 cm)
diameter vacuum pipe. It was found that a quite homogene-
ous mixture of finely ground cement (nominally through 300
10 mesh) and the IGD-101 Microballoons was obtained by this
procedure.
This material was then mixed to form the light-
weight cement slurry. During this, a special ~alliburton
loop densitometer with a range from 8.3 lbs/gal to 10.8
15 lbs/gal (0.99 to 1.29 kg/litre) was employed, which had
been calibrated with water at 8.34 ppg (1 kg/litre). The
mixing was done with a jet mixer and a pump truck equipped
with two Halliburton T-10 pumps, i.e., typical oilfield
equipment. The slurry was mixed at different feed water
20 pressures ranging from 125 to 475 psi (8.79 to 33.4
kg/cm2). The slurry was best mixed at a feed water pres-
sure of 475 psi (33.4 kg/cm2). Then, as mentioned above,
mixed at a rate of 4 barrels per minute (636 litre/min.),
the slurry was quite adequately pumpable, although it
25 appeared somewhat thick. The centrifuge showed the water
content to be in the range of 80% to 85%, which was the
objective; a pycnometer showed slurry specific gravity of
the range of 1.03 to 1.14. The loop densitometer chart
showed ranges of specific gravity from 1.04 to 1.06 which
30 upon calibration against the pycnometer measurements
showed the densitometer to be substantially accurate.
The field processing of the lightweight pumpable
hydraulic cemen-t slurry as has been described is straight-
forward and essentially is as follows: the homogeneously
35 blended dry materials, namely spheres and cement, are
mixed at the well with the amount of water equivalent to
(a) the customary amoun-t of water (API normal to minimum
water contents) plus (b) the excess based on the concen-
tration of spheres, as described above. This produces a
,
slurry with consistency in the range from a maximum con-
sistency of 30 Bc (API) to one with a free water not
exceeding 2 volume percent (~PI) (tests for consistency in
Beardens and for percent free water are as specified in
5 API code RP 10B). When so prepared, the material is
pumped by ordinary cementing trucks Ising the usual
cementing procedures. In one such procedure J the cement
is pumped downwardly through the casing then flows up and
around it, permitting casing manipulation such as scratch-
10 ing, rotating, oscillating, etc., to displace the drillingmud and do a good cement job.
After the cement placement~ we find that ordi-
nary waiting on cement (WOC) times result. The develop-
ment of strength as the slurry sets depends not only on
15 temperature and pressure but on cement class, water con-
tent and presence of additives. Oil well cementing slur-
ries of this invention develop relatively high compressive
strengths within 12 hours as illustrated in Figure l,
whereas previously used low density slurries developed
20 only low strengths after 24 hours. This is particularly
true for slurry densities in the range of about 1.08 to
about 1.4 kg/litre (about 9 to about 12 pounds per gal-
lon).
It is extremely time consuming to determine the
25 reaction of cements to the presence of all types of cement
additives. However, tests to date do not indicate any
difference in the use of the cement additives with these
particular slurries, compared to those prepared in the
absence of the finely divided hollow spheres. In other
30 words, such things as dispersants, accelerators, and the
like operate essentially in the same fashion as before.
It should be emphasized that the cured or set
cement (the solid resulting from the slurry already des-
cribed) has an additional valuable property in addition to
35 high mechanical strengths and low densities. This is that
it has considerably less thermal conductivity than ordi-
nary cements cured from slurries which do not contain
these hollow spheres. As is the case for other heat insu-
lators, the presence of the inclosed gas in the very large
number of hollow spheres incorporated into the set solid
gives a marked decrease in thermal condwctivity. This is
important when setting cement against permafrost or where
well fluids are quite hot.
Two casing strings have been experimentally com-
pleted with the lightweight cement of the present inven-
tion. In both completions, circwlation of cement to the
mudline was observed with a remote camera. These casing
strings were the 24 inch (610 mm) and 16 inch (406 mm)
10 casing strings which, respectively, extended 600 and 1,000
feet (183 and 305 m) below the mud line in 1,000 feet
(305 m) of water. Previous attempts on similar wells to
circulate commercially available low density cements to
the mud line had failed.
The 24 inch (610 mm) casing string was cemented
with about 2,456 cubic feet (70 m3) of slurry formulated
by mixing about 900 sacks (84,600 pounds, 38370 kg) of API
Class A Portland cement with about 1,300 pounds (590 kg)
of calcium chloride, about 144 (65 kg) pounds of defoamer,
20 about 22,500 pounds (10,206 kg) of B37/2000 hollow glass
microspheres and sufficient fresh water to produce a 9.5
ppg (1.14 kg/litre) slurry. The 16 inch (406 mm) casing
string was cemented with about 2,188 cubic feet (62 m3) of
slurry formulated by mixing about 800 sacks (75,200 pounds
25 34110 kg) of API Class A portland cement with about 1,200
pounds (544 kg) of calcium chloride, about 128(58 kg~
pounds of defoamer, about 20,000 pounds (9072 kg) of
B37/2000 hollow glass microspheres and sufficient fresh
water to produce a 9.5 ppg (1.14 kg/litre) slurry. The
30 microspheres, defoamer, and calcium chloride were dry
blended with the cement and then the dry blended mixture
was mixed with the water just prior to cementing the
casing strings.
It will be readily apparent to those skilled in
35 the art of cementing oil wells and the like that the util-
ity of the slurry, of the described method, and of the
resulting set composition are not dependent on the embodi-
ments shown and that a considerable variation can be per-
mitted, as set out in the scope of the appended claims.
