Professional Documents
Culture Documents
SPE 25324
INTRODUCTION
The well cementation design process includes
slurry design, spacer design, hole conditioning
program design, casing attachments selection, and
placement procedure design.
This paper is
concerned with slurry design and more
particularly, the effects of deviations in mixed
slurry density from the design point. Concern
over these effects has been increasing in recent
years. In 1987 an API Task Group in the Eastern
Hemisphere Cementing formed a work group to
study "the effect of poor density control on cement
This group also
slurry properties..... ".'
investigated measurement and control of density.
There were wide differences of opinion as to how
well density was being controlled in the field and
also what was necessary. Additional data and
analysis will be presented which will help clarify
some of the density slurry property relationships
and therefore provide additional means to make
rational decisions relative to density control
specifications.
Density as a slurry property is significant to
wellbore hydrostatic pressure.
However,
deviations from design density may cause changes
in all other slurry properties which mayor may
not cause future well problems. The relationships
defining slurry property dependence on slurry
density are important when establishing meaningful
density control limits. However, closer limits
117
SPE 25324
SPE 25324
y~r5
SPE 25324
* p*f,
Laminar:
[P2).8388
RatiO of lip= PI
(e(P2-8.33).1612
* -(PI-8.33)
-e
1.
Density changes will not affect friction
pressures significantly when in turbulent flow with
non-bentonite slurries
2.
Density changes will affect friction
pressures significantly (up to 35% for a 0.3 lb/gal
density change) when in laminar flow. Thus,
operators should be aware of the fact that density
variations cause much higher friction pressure
changes when in laminar flow as compared to
turbulent flow.
The ratio of hydrostatic pressure due to a density
increase = P2/PI' Using the same density change
as in the above example, the ratio is 1.019 or a
1.9% increase. Thus, density deviations have a
more significant effect on friction pressure than
hydrostatic pressure.
Similar calculations can be made for Reynolds
Numbers using the same assumptions used above.
The ratio of Reynolds numbers for two different
operating densities is:
SPE 25324
FLUID LOSS
Lower fluid loss values will result from higher
mix densities with the same blend (lowering the
mix water/cement ratio will yield lower fluid loss
values).
Figure 6 illustrates typical data for fluid loss as a
function of density variation. Table 2 shows the
rate of change of thickening time per 0.1 Ib/gal
change in mix density. The average percent
change per 0.1 Ib/gal for slurries" A", "C", "0",
and "E" was 4.85%. The highest value was
7.8 %. Thus, it appears that fluid loss is not
dramatically affected by density errors. It appears
that the percent change in fluid loss follows
linearly with the percent changes in mix water but
too little data has been examined to make
generalizations. Considering the effect of a 0.3
Ib/gal density error on the tested slurries, it can be
seen that fluid loss may vary as little as 3.9% with
slurry "0" to a high of 23.4% with Slurry "C".
The variation in Slurry "C" could possibly cause
an excessive fluid loss, particularly if a greater
error in density occurred, such as a 0.5 Ib/gal.
THICKENING TIME
Lower thickening time values will result from
higher mix densities with the same blend (lower
amounts of mix water will yield shorter thickening
times).
Figure 8 illustrates typical data for
thickening time as a function of density variation.
Table 2 shows the rate of change of thickening
time per 0.1 Ib/gal change in mix density. The
average percent change per 0.1 lb/gal for slurries
"A", "B", "C", and "E" was 3.9%. Thus, even
with a mixing density error of + 0.3 Ib/gal, only
a 11.7 % error in the expected thickening time
would occur.
Variances of this order of
magnitude are not significant. Note, however that
in Slurry "0" there was a 35% change per 0.1
Ib/gal. This is very significant. What actually
happened was that the lowest density, 12.3 Ib/gal,
became so diluted with water that the thickening
time exceeded 18 hrs.
This illustrates an
important point -- that specific knowledge about
the particular slurry being mixed is necessary.
Most slurries will not have a significant thickening
time change with small density variations, but
some might. It would be desirable to select slurry
designs which exhibit characteristics more like
Slurries "A", "B", "C", and "E", thus helping
avoid potential problem jobs.
FREE WATER
Slurry free-water values decrease with increasingly
higher densities, (decreasing the mix water/cement
ratio will decrease the amount of free water).
Figure 7 illustrates typical data for free water as a
function of density deviations. Table 2 shows the
rate of change of free water per 0.1 Ib/gal change
in mix density. These values vary from 0 for
Slurry "C" to 12.8% for Slurry "0". This data
seems quite irregular but there is an explanation
behind these variations.
Slurry "C" had
essentially zero free water and therefore percent
change was meaningless. Slurries "A" and "E"
were more typical of what might be expected.
Slurry "B" had less than 1% free water at the
lower density and thus a small decrease in free
water translated into a high percent change.
Slurry "0" was a case of initially having too much
SPE 25324
CONCLUSIONS
Very limited laboratory data shows how individual
slurry characteristics vary with slurry density.
Means for estimating compressive strength and
slurry viscosity deviations with density have been
presented. Slurry parameters of fluid loss, free
water, gel strength, and thickening time have been
investigated for selected compositions. Percent
change in these parameters per 0.1 lb/gal density
deviation have been calculated. The magnitude of
these changes were generally small and would not
be cause for alarm. However, slurries may be
close to a limit where a significant change in a
property could occur.
Therefore, it is
recommended that knowledge of the effects of
density deviation be known to avoid problems
during the cementing process. Tests should be
conducted for general classes of designs and when
new products are being contemplated.
1.
2.
3.
4.
TABLE 1
Slurry
Identification
Mean
Density, lb/gal
-SlurriGeneral Description
A
B
C
D
E
F
G
H
16.46
14.56
16.5
12.8
17.3
12.5
15.8
122
Table 2
7000
FJm
Free
Water Loss %
%
ThIckenmg
Time
%
C'-l
Crulll ~
Mean
Density,
lb/gal
6000
~
5000
'iii
Q.
4000
~ 1'72 """
.........-::::
0.
i 3000
I~
"-
iii
N
.J:i.
'24 ......
1% Ir utA. SI'....
lIE
CrYltlSlf'....
2000
1000
-1
-0.5
0.5
Slurry .. E..
Density Deviation, Ib/ gal
Tcrget Density: 17.3 1b/ga1
tT1
>
Z
I\)
c.J
CHG COMP
12
V)
STR
CHG
"r1
VISCOSITY
i 12
~,
11
11
0-
Curve "A"
10
10
I'
0Z
..
a:
l-
V)
::::
'.
Slurry "c"
'~
'.
V)
V)
VISCOSITY
-; 9
0- GEl-VARIOUS :t:
- AOOITIVE-VARIOUS :t:
~8
X - ow WIFLUID LOSS AOOITIVE
+ - oW WIlIEIGHT AOOITIVE
--l 7
..
+,
a:
'
0-
::;
0
'
3
10
11
12
13
14
15
DENSITY, LB/GAL
16
17
18
19
~-
C'-l
>
Z
o
en
~~~~
4
o
.. .
. .._-
-+----+---""
11
12
13
14
15
16
17
18
DENSITY. LB/GAL
Figure 3: Percent Change in Property with 0.1 lb/gal of Density
4
'3
19
-..l
280
500
00
240
i..
400 L-
0200
0
~ 160
,_N,
.....
iii
~
a.
u
.00
II~b/"'l
f-
... 120
fJ)
Vi
...c
fJ)
200
80
>
Cl>
J:
VI
40
o-f': l100
Slurry "E"
100 L-
200
300
400
500
600
a
10
Figure 4:
Jf
,
>-'
"
"
'~
"
11
SLURRY "G"
4>
13
14
DENS I TY,
SLURRY "0"
~l?
,.
"
17
LSI GAL
~SLURRY "E"
,.
>-<:
"
+ SLURRY "H"
Figure 5: Viscosity Correlation - Class "A" with 2% Light Weight Additive in Sea
Water
......
I\)
-l:>-
50
E40
iii
230
T
~2
tj
.:Ql
ii
CI)
tT.I
.~
Ql
Ql
...
20
CI)
i:l
g
~
60
..,)
u..
10
-1
-0.5
Slurry .. E ..
0.5
Figure 6:
-1
-0.5
Slurry "E"
0.5
Figure 7:
CI)
VI
CI:l
12
10
"
....
,/s.
Tw
WliGlSeI
.~
I-
50
tv
g 40
",Q
.30
.I:
0,
~ 20
-1
-0.5
.-------.J
r---T. T.
'-----~
SlJrry .. E ..
Density Deviation,
Target Density: 17.31b/gal
Figure 8:
0.5
Ib/ gal
,J:o.
IOIf'ftlM~
l,....---""
+
4
~
I\:l
01
~
W
I ~k'~ 1
:I
.I:
60
---
,........
~
--------:
10 mlrU,
I ........
iii
Q;
D 10
-1
-0.5
Figure 9:
0.5
>
z
>
"!j
t"'"
CI:l
>
CI:l
\0