Smart Cementing Drilling Muds Monitoring Deepwater Wellbore Enhancement

Smart Cementing Materials and Drilling
Muds for Real Time Monitoring of

Deepwater Wellbore Enhancement
RPSEA Ultra-Deepwater Program
Final Report Small Model Tests
Project No. 10121-4501-01
C. Vipulanandan and R. Krishnamoorti
Center for Innovative Grouting Materials and

Technology (CIGMAT)
University of Houston
December 2014
Project 10121-4501-01 / UH-CIGMAT
Final Report-Small Model Tests
Table of Contents
Page
Table of Contents.................................................................................................................... 2
Abstract.......................................................................................................
1. Introduction....................................................................................................
2. Objectives.......................................................................................................
3. Theory and Concept................................................................................................... 7

4. Small Model Tests..........................................................................................................
10
5. Conclusions...... 38
6. Bibliography ..
39
ABSTRACT
In this study, small models were designed, built and used to demonstrate the concept of
real time monitoring of the flow of smart drilling mud, space fluid and smart cement and
hardening of the cement in place. Also a new method has been developed to measure the
electrical resistivity of the materials using the two probe method. Using the new concept, it has
been proven that resistivity dominates the behavior of drilling mud and smart cement. LCR
meters (measures the inductance (L), capacitance (C) and resistance (R)) were used at 300 kHz
frequency to measure the changes in resistance. Several laboratory scale model tests have been
performed using instrumented casing with wires and thermo couples. When the drilling mud was
in the model borehole the measured resistance was the highest based on the high resistivity of the
drilling mud. Notable reduction in electrical resistance was observed with the flow of spacer
fluid and cement. Change in the resistance of hardened cement has been continuously monitored
up to about 100 days. Also a method to predict the changes in electrical resistance of the
hardening cement outside the casing (Electrical Resistance Model - ERM) with time has been
developed. The ERM predicted the changes in the electrical resistances of the hardening cement
outside the cemented casing very well.
1. INTRODUCTION
As deep water exploration and production of oil and gas expands around the world, there
are unique challenges in monitoring the well construction beginning at the seafloor. Two
separate studies performed on oil well blowouts, one done between the years of 1971 to 1991
and the other study was done during the period of 1992 to 2006, before the deepwater horizon
blowout in the Gulf of Mexico in 2010. The two studies clearly identified cement failures as the
major cause for blowouts [Izod et al. 2007]. Cementing failures increased significantly during
the second period of study when 18 of the 39 blowouts were due to cementing problems [Izod et
al. 2007]. Also the deep-water horizon blowout in the Gulf of Mexico in 2010, where there was
eleven fatalities, was due to cementing issues [Carter et al. 2014]. With some of the reported
failures and growing interest of environmental and economic concerns in the oil and gas
industry, integrity of the cement sheath is of major importance. At present there is no technology
available to monitor the cementing operation real time from the time of placement through the
entire service life of the borehole. Also there is no reliable method to determine the length of the
competent cement supporting the casing.
1.1 OIL WELL CEMENT
Oil well cementing is done to provide a protective seal to the casing, prevent lost
circulation and blowout and to promote zonal isolation. The standards of API suggest the
chemical requirements determined by ASTM procedures and physical requirements determined
in accordance with procedures outlined in API RP 10B and ASTM. There are several classes of
cements which could be used for oil well cementing. The main focus of this study was to make
the cement more sensing to the changes observed in the cement sheath.
1.1.1 Modifiers: Several modifiers (additives) have been used in the cement slurry to modify the
sensing properties. The modifiers used in study include silicates, meta kaolin, fly ash and
nanoparticles. The study has resulted in number of publications (Vipulanandan et al. 2014a and
2014b)
1.1.1a Sodium Meta Silicates
A series of experiments evaluated the smart well cement behavior with and without up to
0.3% sodium metasilicate (SMS) to determine the sensitivity of the electrical resistivity of
cement from curing to hardened state was investigated. The test results showed that SMS
reduced the electrical resistivity of the water and cement slurries based on the amount. The SMS
up to 0.3% also affected the rheological properties, setting characteristics, and the piezoresistive
properties of the smart cement. In a 24-hour period the maximum change in the electrical
resistivity (RI24) for the cement without SMS (0.4 water-to-cement ratio) was 175%. The RI24 for
the cement with SMS varied with the amount of SMS. Addition of 0.2% SMS had a minimal
effect on the compressive strength. The smart cement was piezoresistive with the addition of
SMS but the sensitivity was decreased.
pH
Addition of SMS to the water increased the pH. With the addition of 0.1% SMS, the
water pH increased from 7.7 to 11.8, a 50% change in the pH. With the addition of 0.3% SMS
the pH was 12.4.
Resistivity
The resistivity of sodium metasilicate solution was determined with the conductivity probe.
SMS solution was very sensitive to electrical resistivity. The resistivity of water decreased from
21 .m to 4.15 .m with an addition of only 0.1% SMS, a 80% reduction in resistivity. The
addition of more SMS further reduced the resistivity of the solution. The following relationship
is proposed based on the experimental results:
= o S/(E +DS) .(1)
where:
= resistivity of the sodium metasilicate solution
o = resistivity of tap water without sodium metasilicate (21 .m)
S = Concentration of sodium metasilicate (% by weight)
Parameters E and D are model parameters: parameter E represent the initial rate of change and
parameter D determines the ultimate resistivity. Experimental results matched very well with the
proposed model with a coefficient of determination (R2) of 0.98, and parameters E and D were
found as 0.0016/.m and 0.047/.m, respectively.
Rheological properties
Gel Strength
For the smart cement, after 10 seconds gel strength was 12 lb/100 ft2, and after 10 minutes gel
strength was 14 lb/100 ft2.With the addition of 0.2% SMS to the cement the 10-second gel
strength was 15 lb/100 ft2 and the 10-minute gel strength was 17 lb/100 ft2. Hence, addition of
0.2 SMS increased the 10-second gel strength by 25% and the 10-minute gel strength by 21%.
Effect of SMS
It was evident from the rheological tests on cement slurries with and without 0.2% SMS
that the addition of SMS negatively affects the shear-thinning behavior of cement. Addition of
SMS increased the viscous behavior of the cement. For instance, the viscosity of cement without
SMS at a shear strain rate of 100 (1/sec) was 146 cP. At the same shear strain rate the 0.2 percent
SMS sample had viscosity of 225 cP, a 54% increase in viscosity.
Modeling
Due to the shear-thinning behavior of the cement slurries the Bingham plastic model was not
an accurate model to estimate the shear strain rate shear stress relationship. To predict the shear
strain rate - shear stress relationship, the Herschel-Bulkley model (Eqn.2) and hyperbolic model
(Eqn. 3) were used to predict the experimental data.
The HerschelBulkley (H-B) model developed in year 1926 is as follows:
= o + k n ..(2)
where:
= shear stress
= shear rate
o= yield stress
k and n are model parameters.
From the smart cement, the k and n were found to be1.09 and 0.78, respectively. The
coefficient of determination (R2) was 0.97. The yield stress was found as 1.87 lb/100 ft2.
For the cement slurry with 0.2% sodium meta-silicate the k and n (for HerschelBulkley
model)were found to be 0.61 and 0.88, respectively. The coefficient of correlation was 0.99. The
yield stress was 13.8 lb/100 ft2, a notable increase in the yield stress with the addition of 0.2%
SMS.
The new hyperbolic model developed by Vipulanandan et al. (2014) is as follows:
= o + /(A+B) (3)
where.
= shear stress
o= yield stress
A and B are model parameters.
Here for the smart cement parameters A and B were found as 2.83(Pa.sec)-1and 0.001Pa-1,
respectively, and the coefficient of determination (R2) was 0.97. The yield stress was 5.01 lb/100
ft2. The apparent viscosity at 600 rpm was 106 cP.
For the smart cement with 0.2 SMS, the hyperbolic model parameters A and B were found
as3.21(Pa.sec)-1 and 0.0004 Pa-1,respectively, and the coefficient of determination (R2) was
found to be 0.99. The yield stress was17.4 lb/100 ft2.With the addition of 0.2% SMS the apparent
viscosity at 600 rpm was 136 cP, a 28% increase.
1.1.2 PIEZORESISTIVE BEHAVIOR
Several methods of modifying the cement to be more piezoresistive are being
investigated.
Carbon Fiber: After 7 days of air curing, the specimen (w/c = 0.4) with 0.075 percent carbon
content (total weight of cement) was tested under compression loading. Typical change in
resistivity with compressive stress is shown in Fig. 1 (DC measurement). The change in
resistivity in the bulk material was much higher than the strain response. For example, the
change in bulk resistivity was 45 percent as compared to a strain of 0.05 percent, at a
compressive stress of 750 psi. This shows the magnification of the resistivity response of the
modified cement with 0.075 percent of carbon fibers. The axial compressive strain was about 0.2
percent at failure for the fiber-modified cement and the resistivity change was 80 percent. The
resistivity change was about 400 times higher than the change in strain.
When the AC measurement (f = 300 kHz) was used it was possible to overcome the
contact resistance and the Piezoresistive response was substantially increased as shown in Fig. 2.
Crack
CF - 0.075%
W/C - 0.4
Curing time - 7 days
Figure 1. Piezoresistivity Behavior of Oil Well Cement with 0.075% CF
1.2 Drilling Muds

The study focused on enhancing the sensing properties of water based and synthetic
based drilling fluids. The rheological properties of the drilling fluids have been related to the
electrical resistivity, selected to be the sensing property.
Crack
Figure 2. Piezoresistivity Behavior of Oil Well Cement with 0.075% CF (AC frequency =300 kHz)
2. OBJECTIVES
The overall objective of the small model study was to demonstrate the use of the smart
drilling mud and smart cement slurry (enhance sensing properties) that can be integrated with
real-time monitoring of the operations for improving the wellbore operation and long-term
integrity.
The objective for the small model study is as follows:
(1) demonstrate with small models (simulating the well cementing) the real-time monitoring of
smart drilling mud (SDM) and smart cement slurry (SCS) during installation and service life of
the well.
3. THEORY AND CONCEPTS
It was very critical to identify the sensing properties for the cement and drilling mud that
can be used to monitor the performance. After years of studies and based on the current study on
oil well cements and drilling muds, electrical resistivity () was selected as the sensing property
for both cements and drilling muds. This makes it unique since same monitoring system can be
used to evaluate the performance of the cement and drilling muds. Hence two parameters
(resistivity and change in resistivity) will be used to quantify the sensing properties as follows:
R = (L/A) = K ..(4)
Where R = electrical resistance
L = Linear distance between the electrical resistance measuring points
A = effective cross sectional area

K = Calibration parameter is determined based on the resistance measurement
method
Normalized change in resistivity with the changing conditions can be represented as follows:
/ = R/R ................... (5)
In resistivity of the materials () to changes (composition, curing, stress, fluid loss and
temperature) has been quantified. Correlating the changes such as composition, curing, stress,
cracking, fluid loss and temperature to the resistivity () (Eqn. (4)) and change in resistivity ()
(Eqn. (5)) will support the monitoring of the material (cement and drilling mud/fluid) behavior.
IMPEDANCE SPECTROSCOPY MODEL (Vipulanandan et al. 2013)
Equivalent Circuit
Identification of the most appropriate equivalent circuit to represent the electrical
properties of a material is essential to further understand its properties. In this study, an
equivalent circuit to represent the Smart Cement and Smart Drilling Mud was required for better
characterization through the analyses of the IS data. There were many difficulties associated with
choosing a correct equivalent circuit. It was necessary somehow to make a link between the
different elements in the circuit and the different regions in the impedance data of the
corresponding sample. Given the difficulties and uncertainties, researches tend to use a
pragmatic approach and adopt a circuit which they believe to be most appropriate from their
knowledge of the expected behavior of the material under study, and demonstrate that the results
are consistant with the circuit used.
In this study, different possible equivalent circuits were analyzed to find an appropriate
equivalent circuit to represent the smart cement and drilling mud.
Case 1: General Bulk Material Resistance and Capacitor
In the equivalent circuit for Case1, the contacts were connected in series, and both the
contacts and the bulk material were represented using a capacitor and a resistor connected in
parallel (Fig. 3).
In the equivalent circuit for Case 1, Rb and Cb are resistance and capacitance of the bulk
material, respectively and Rc and Cc are resistance and capacitance of the contacts, respectively.
Both contacts are represented with the same resistance (Rc) and capacitance (Cc) as they are
identical.Total impedance of the equivalent circuit for Case 1 (Z1) can be represented as follows:
Z1 ( )
2Rc2Cc ( ) Rb2Cb ( )
Rb ( )
2 Rc ( )
2 2 2
2 2 2
1 2 Rb2Cb2 1 2 Rc2Cc2
1 Rc Cc 1 Rb Cb
(6)
where is the angular frequency of the applied signal. When the frequency of the applied signal
was very low, 0, Z1 = Rb + 2Rc, and when it is very high, , Z1= 0.
Case 2: Special Bulk Material - Resistance Only
In Case 2, as a special case of Case 1, the capacitance of the bulk material (Cb) was
assumed to be negligible (Fig. 4).
The total impedance of the equivalent circuit for Case 2 (Z2) is as follows:
Z 2 ( ) Rb ( )
2 Rc ( )
2Rc2Cc ( )
j
.
1 2 Rc2Cc2
1 2 Rc2Cc2
(7)
When the frequency of the applied signal was very low, 0, Z2 = Rb + 2Rc, and when it is
very high, , Z2 = Rb (Fig. 5).
Cc
Cb
Cc
Rc
Rb
Rc
Figure 3. Equivalent circuit for Case1
Cc
Cc
Rb
Rc
Rc
Figure 4. Equivalent circuit for Case 2
Impedance
Case 2
Case 1
0
0.1
10
100
1000
10000
100000
1000000
Frequency (Hz)
Figure 5. Comparison of typical responses of equivalent circuits for Case 1 and Case 2
Testing of smart cement and smart drilling mud clearly indicated that Case 2 represented their
behavior at a frequency of 300 kHz.
4. SMALL MODEL STUDY
Laboratory models were designed and built at the University of Houston. The model was
built to monitor the slurry level (drilling mud, spacer fluid and cement) during the installation
and hardening of the cement. The observed resistance with time clearly indicated the level of
slurry and to determine the depth at which the drilling fluid and cement was located. Several
models were built separately and tested separately to demonstrate the real-time monitoring
concept.
4. 1. Model #1 (Water based drilling mud and cement slurry with w/c ratio of 0.8)
Materials and Method
The model was built using a plexiglass and metal pipe as shown in the Fig. 6 to simulate
the formation and casing. The casing was instrumented with electrical wires to monitor the
resistance change. The distance between two sensors was 4 inches and there were six levels of
sensors as shown in the Figure 1. Different combinations of the sensors were connected to a 300
Hz LCR device to measure resistance between those sensors.
10
LCR device
Plan view
Transparent
mold
Metal casing
Sensors
Circuit board
Wire outlets
Stand
Figure 6. Laboratory scale oil well model and monitoring system
The horizontal electrical wire leads (sensors/monitors) are noted as a, b, c or d. The vertical wire
leads are marked by 1 to 6. Figure 7 shows the different levels of liquid in the simulated well.
Figure 7: Different levels of cement in the oil well model

Experimental Setup
Initially water was placed in the model to represent drilling mud. Using a tube, around 26
Liter of water was poured into the well model and allowed to fill from the bottom of the model.
The electrical resistance was measured with height when water reached each level. After
11
measuring the resistance for all six levels water was removed and study was continued with
cement slurry (Fig. 7).
Same amount of oil well cement class H slurry was used with a water to cement ratio 0.8.
The cement was modified with an addition of 0.075% conductive fiber (CF) by total weight. The
same procedure was followed as for water and resistance measurements were taken when the
slurry reach each levels.
Results and Discussion
(a) Monitoring the Drilling Mud (Water) Level
Resistance measured for different vertical levels of well is shown in Fig. 8. When there
was no water in the well (at 0 level of water) the resistance was in the range of 450 to 650 k
which can be considered as the air resistance for the particular distance monitored at the relative
humidity of the lab. When water level comes to level 1, all the vertical resistances dropped down
to 80 to 120 k ranges. This sudden change clearly showed that water reached level 1.
Figure 8: Vertical resistance measurement in the oil small model with changing water level
12
When water reached level 2, all vertical resistance combinations observed a small
reduction in their values. But sensors c1-c2 (resistance in section-c and vertical line between
level 1 and 2) observed drastic change in the resistance. It went down to about 1000 from
about 95,000 . It happened because the earlier air resistance was replaced by water resistance
when water fills the space between level 1 and 2 while the others levels experienced air
resistance.
When water reached level 3, resistance between c1-c3 reduced drastically. The same
pattern was observed for other sets of readings too, as shown in the Fig. 8. This consistent
behavior showed that the level of the drilling fluid can be monitored effectively by measuring the
resistance. The continuous reduction in the resistance was observed in the model with the rising
water level. When the water level went upwards, air was replaced by water gradually. Water has
relatively lesser resistance than air. Water and air can be considered here as connected in series.
So the resulting resistance reduced with rising water level. And the drastic change was observed
when the circuit was completed with water, without any air.
Figure 9. Horizontal resistance measurements with varying heights of water level
13
Figure 9 shows the changes in the resistance for horizontal level of sensing points. When
water reached each horizontal level, the resistance dropped from several hundred k to few
ohms. This sudden change in the resistance indicated that the level of the water reached up to
that particular height.
Figure 10 shows how the resistance changed with different levels of water for individual
horizontal levels monitored. When the water level reached each level, resistance for that level
reduced suddenly. For example, resistance between a3 and c3 (at horizontal level 3) reduced
drastically when the water reached level 3.
Figure 10. Variation of resistance with rising water level

(b) Monitoring the Cement Slurry Level
The same procedure was used to demonstrate the smart cement slurry rise and the vertical
resistances are reported in Figure 11. Similar pattern to water rise was observed which enabled
the sensing method to detect the level of the cement slurry.
14
In contrast to 1000 resistance of the drilling mud (water) the rage of resistance of
cement slurry was in the range of 20 to 50 for the time interval monitored. This observation
showed that the material can be distinguished based on the resistance value while identifying the
level of the slurry.
Figure 11: Variation of vertical resistance with different level of cement slurry
With time when the cement started to set, due to bleeding in the cement slurry free water
accumulated above the cement. This effect was demonstrated by using the cement slurry with a
water-to-cement ratio of 0.8. As shown in the Fig. 12, the resistance measurements clearly
distinguished between free water and cement. At the earlier part of the experiments the water
resistance was found to be in the range of 1000 . But the free water resistance value was in the
range of 30 to 40 because of the mixed ions from the cement slurry. This observation showed
that, using these resistance measurements, free water also can be detected.
Horizontal resistance measurements for cement slurry are reported in Fig. 13, which
showed similar pattern to the drilling mud (water). This observation again confirmed that the
cement slurry level can be detected using resistance measurement.
15
Figure 12: Resistance for free water and cement slurry
Figure 13. Horizontal slurry resistance with slurry level

16
Figure 14, shows the resistance variation with rising smart cement slurry level of each
level of sensors which is again similar to that of drilling mud.
Figure 14: Horizontal resistance variation of cement slurry with slurry level
17
4. 2. Model#2 (Oil based drilling mud, Spacer Fluid and cement slurry with w/c ratio of
0.38 )
Materials and Method
A similar model (3.5 ft. height and 6 inches in diameter) with a 2 inch diameter casing,
similar to the previous model was built with seven levels of sensors (wire leads). Again the
distance between two sensors (wire leads) was 4 inches.
Oil based mud was prepared by mixing mineral oil, water and cetyltrimethylammonium
bromide (CTAB) surfactant together. The ratio of oil to water ratio was 4:1. Initially, 1% of
cetyltrimethylammonium bromide surfactant was mixed slowly with water and then oil was
added gradually to the surfactant solution. The density of the oil based mud was 7.2 ppg and the
electrical resistivity was 110 .m.
The spacer fluid was prepared by mixing 0.5% Guar gum, 0.4% bio-Surfactant and 3%
KCl. The density of the spacer fluid was about 8.3 ppg (1 g/cc) and the electrical resistivity was
0.8 .m. Oil well cement class H was used with a water-to-cement ratio 0.38. The cement was
modified with an addition of 0.075% carbon fiber by the total weight of the cement slurry.
The horizontal electrical wire leads (sensors) are referred to as a, b, c or d. The vertical
wire leads (sensors) are referred to as level 1 through 7 (Fig. 6). Different combinations of the
sensors were connected to a 300 kHz LCR device to measure resistance between those wire leads
(sensors).
Experimental Setup
Initially about 10 L of drilling mud was placed in the annulus of the well model from the
bottom of the model towards upward using a pressure chamber as shown in the Fig. 15. The
electrical resistances of difference combinations of sensors were measured for each levels of
mud. After measuring the resistance for all seven levels drilling mud was replaced by spacer
fluid of 10 L gradually by applying pressure around 20 psi from the pressure chamber (Fig. 15).
The same procedure was followed for spacer fluid as of drilling mud to measure the resistance
when the spacer fluid reached each level.
After measuring the resistance of spacer for all seven levels, spacer was replaced by the
cement slurry by applying pressure of about 20 psi from the pressure chamber and same
procedure was followed as before to continuously monitor the electrical resistances between the
selected wire leads.
18
Pressurizing chamber
Pressure gauge
Cement
Outlet tube
Inlet tube
Spacer
Cement
Figure 15. Experimental setup of Small Model #2 where cement is displacing spacer fluid
Results and Discussion
(a) Monitoring the Drilling Fluid
Resistance was measured for different vertical levels in the small well model as shown in
Figure 16 with time. When there was no fluid in the well (at 0 level of the casing) the resistance
was in the range of 155 to 205 k which represented the air resistance for the particular distance
monitored at the relative humidity of the laboratory. When oil based mud level came up to level
1, all the vertical resistances started to reduce to 35 to 60 k range. This sudden change clearly
showed that oil based mud reached level 1. Similar reductions in electrical resistances were
observed with horizontal measurements.
When oil based mud reached level 2 (observed through the plexiglass model), the sensors
a1-a2 (resistance in the vertical direction between level 1 and 2) showed sudden change in the
electrical resistance. It resistance reduced to 14 k from about 36 k. The change happened
because the drilling mud the drilling mud with the resistivity of 110 .m filled the space the
between level 1 and 2. When the mud reached level 3 (observed through the plexiglass model),
the resistance between a1-a3 dropped (Fig. 16). The same pattern was observed for the other set
of readings too. This consistent behavior showed that the level of the drilling fluid can be
monitored effectively by measuring the resistance. The electrical resistance changes observed
during the placement of the drilling fluid was very similar to the first model.
19
Figure 16: Vertical resistance measurements for the oil based drilling fluid
(b) Monitoring the Spacer Fluid
The same procedure that was used to monitor the drilling fluid was followed for the
spacer fluid. Figure 17 shows the changes in the resistance readings for the spacer fluid. When
the space fluid filled the space the resistance was rapidly lowered because of the lower resistivity
of the space fluid.
Figure 17. Vertical resistance measurements for spacer fluid

20
(c) Monitoring the Cement Slurry

After filling up model totally with the spacer fluid, spacer was displaced by cement
slurry. The same procedure was followed to place the cement slurry and to monitor the changes
in the resistances. Figure 18 shows the resistance readings for cement slurry.
Figure 18. Vertical resistance measurements for cement slurry

When there was no cement slurry in the well (at 0 level of water) the resistance was in the
range of 25 to 35 representing the presence of the spacer. When cement slurry reached level 1,
all the vertical resistances increased to between 55 to 67 ranges. This change clearly showed
that cement slurry reached level 1. The increase happened because the resistivity of the cement
was higher than that of spacer fluid. In the process of displacing the spacer fluid, cement got
contaminated with the spacer fluid and therefore the resistance values dropped to 30 to 50
between level 1 and 2. Based on the increase in the resistance, no contamination happened at
level 1. Also because of the hydration process of the cement, the resistance increased with time
because the resistivity of the cement increased with time.
(d) Continuous Monitoring the Fluids used in the Model
Figure 19 shows the resistance change with time in the well for different types of the
fluids used. Initially the air resistance dropped to oil based drilling muds resistance followed by
resistance representing the spacer fluid and cement slurry.
21
Figure 19. Vertical resistance measurements in the oil well model

The variation of resistance with time clearly showed that the level of the slurry can be
monitored by measuring the resistance with time.
Figure 20(a) shows the changes in the electrical resistance of the bottom sensors a1-a2
and Figure 20(b) shows the resistance between the bottom and top sensors a1-a7. In Fig. 20(c)
changes in resistance for both cases are compared. Because the bottom level sensors were filled
with different fluids before the top level sensors, the change of resistance for the bottom level
sensors were out of phase with the top level sensors.
22
Figure 20. Resistance change with time for selected combinations of sensors
4.3 MODEL #3
In this model study the focus was on monitoring the hydration of the hardening cement
sheath supporting the casing. The K parameter (Eqn. (4)) for different wire spacing are shown in
Figures 21 through 23 and the average value, maximum value, and minimum values are
summarized in Tables 1 through 3. For wire setup-a (similar to setup a in Fig. 6), the average K
parameter were found to be varying from 57.9 to 58.6 m-1 with standard deviations varying from
2.0 to 9.1 m-1 for different wire spacing. For wire setup-b (similar to setup b in Fig. 6), the
average K parameter varied from 49.8 to 56.2 m-1 with standard deviations varied from 3.9 to 5.4
m-1 for different wire spacing. And for wire setup-c, the average K parameter varied from 58.7
to 68.1 m-1 and the standard deviations varied from 10.6 to 13.7 m-1 for different wire spacing.
23
Figure 21: Variation of Parameter K for wire setup-a
Table 1: Variations of parameter K for wire setup-a

Wire Spacing
Parameter K
6 inch
12 inch
18 inch
24 inch
Avg. (m-1)
58.0
57.9
57.9
58.6
Min. (m-1)
49.1
44.8
55.2
55.4
Max. (m-1)
67.4
71.4
59.6
61.7
Std. dev. (m-1)
7.4
9.1
2.0
4.5
Figure 22: Variation of parameter K for wire setup-b

24
Table 2: Variations of K parameter for wire setup-b

Wire Spacing
Parameter K
6 inch
12 inch
18 inch
24 inch
Avg (m-1)
49.8
50.3
51.5
56.2
Min. (m-1)
42.6
44.7
43.9
53.4
Max. (m-1)
57.2
59.7
55.4
58.9
Std. dev. (m-1)
5.1
5.4
5.2
3.9
Figure 23. Variation of parameter K parameter for wire setup-c
Table 3: Variations of K parameter for wire setup-c

Wire Spacing
Parameter K
6 inch
12 inch
18 inch
24 inch
Avg. (m-1)
58.7
59.7
64.9
68.1
Min. (m-1)
47.9
49.0
54.0
58.4
Max. (m-1)
78.0
77.4
80.0
77.8
Std. dev. (m-1)
10.6
11.1
11.1
13.7
25
4.1.3.3 Resistivity of the smart cement slurry with curing time: The Resistivity of the cement
slurry with curing time of up to 100 days was determined using a the smart cement slurry
samples (2 inches diameter and 4 inches height cylindrical mold) that was used for the small
model#3 study. The resistivity increased with curing time under the curing of room temperature
and humidity (Figure 24).
Figure 24. Variation of smart cement resistivity with curing time for samples cured under
room conditions (23oC and 50% relative humidity (RH))
4.1.3.4 Predicted and measured resistance for hardening cement
Using the parameters K in Tables 1 through 3 and the resistivity-time relationship (Fig.
24), the changes in the cement sheath resistance in the small model#3 was predicted using the
relationship in Eqn. (4). Figures 25 through Figure 30 show the variations of the predicted
resistance value and also the actual measured values for different wire setup/combinations.
Figure 31 through Figure 33 shows the variations of the predicted horizontal resistance and
actual measured values for setups (a) and (b) at different levels.
(a) Vertical resistance
Wire setup-a
For the wire setup-a, the wire combination a1-a2 showed that the predicted values were
lower than the measured values up to 14 days of curing but after that the measured resistance
values were in the range of the predicted resistance (Figure 25). This may be because in the small
26
model#2 the cement is hydrating under pressure and temperature (Fig. 33) and the resistivity
used (Fig. 24) to predict the resistance was cured under room condition under no pressure.
-a
Figure 25. Comparing the predicted and measured resistance for wire setup-a for wire
combination a1-a2.
For wire setup-a wire combination a1 and a4, the measured values were very close to the
predicted values (Figure 26). The wire at level a4 is very close to the surface of the small model
which showed very similar hydration to the test sample (Fig. 24). The measure resistance values
matched very well with the predicted values.
27
-a
Figure 26. Predicted and measured resistance for wire setup A for wire combination a1-a4
Wire setup-b
For wire setup-b, the wire combination b1 and b3 showed that the predicted values were
lower for about 30 days of curing of curing (Figure 27). As mentioned before the initial
hydration of small model#3 was different (see Fig 33 for temperature increase) from the samples
curing in the 2-inch x 4-inch mold under room condition. After 1 month, the measured values
matched very well with the predicted values since the curing temperatures were very similar (see
Fig. 33).
28
-b
Figure 27. Predicted and measured resistance for wire setup B for wire combination b1 and
b3.
For wire combination b1 and b4, the measured values are very close to the predicted values
(Figure 28). Similar trend was observed with wire a1 and a4 (Fig. 26).
-b
Figure 28. Predicted and measured resistance for wire setup B for wire combination b1 and
b4.
29
Wire setup-c
For wire setup-c, for the wire combination c1 and c2, the initial predicted values were slightly
lower than the measured resistance as observed with setup-a and setup-b. In this region the
cement was curing at a higher temperature and pressure as mentioned before. After two weeks of
curing, the measured resistance matched very well with the predicted resistance (Fig. 29).
Figure 29. Predicted and measured resistance for wire setup C for wire combination c1 and
c2.
And for wire combination c1 and c4, the measured values are within the range of the predicted
values (Figure 30) which was similar to what was observed with setup-a and setup-b.
30
-c
Figure 30. Predicted and measured Resistance for wire setup-c for wire combination c1 and
c4.
(b) Horizontal resistance
For horizontal wire setup a1-b1 at level 1, the predicted values were very close to the
measured values (Figure 31) over a period of 100 days. This prediction was somewhat different
from the resistance predicted vertically between level 1 and 2.
Figure 31. Predicted and measured resistance for horizontal wire setup a1 and b1 at level 1
(bottom)
31
For horizontal wire setup a3-b3 at level 3, the predicted resistance values are also very close to
the measured values (Figure 32).
a3-b3
Figure 32. Predicted and measured resistance for horizontal wire setup a3-b3 at level 3
(middle).
4.1.3.5 Temperature variation during the curing time.
Thermocouples were placed at level 1 with wire setup a (bottom of the model borehole), at level
3 with wire setup b (middle of the borehole), and at level 5 with wire setup c (top of the model
borehole).
At the bottom (level 1), the temperature showed a decreasing trend changing from 34oC to 22 oC
(room temperature) within the first two weeks and then remained at a constant temperature
around room temperature (Figure 33). Hence around level 1 the cement was curing at a higher
temperature compared to the cement samples curing at room temperature.
32
-a1
Figure 31. Temperature variation at level 1 (bottom) in the small model#3

At the middle (level 3) of the borehole, the temperature change showed a decreasing trend
changing from 31oC to 22oC (room temperature) within first two weeks and then remained
constant (Figure 32). The initial temperature was somewhat lower than what was observed at
level 1.
-b3
Figure 34. Temperature variation at level 3 (middle) in the small model#3

33
Also At the top level (level 5) in the borehole, the temperature decreased from 29.5 oC to 20 oC
(room temperature) within first two weeks, then showing a constant temperature of around room
temperature (Figure 35).
-c5
Figure 35. Temperature variation at level 5 (top) of the borehole model
34
4.4 Pressure test
2 Smart
cement
Pi
2 Steel
casing Top view
(b) Plan View
(a) Elevation View
Figure 16. The configuration of the pressure applied in the inside the steel casing
Air pressure (Pi) was applied inside the casing to load the cement-sheath and the
electrical resistance (Ro in Ohms) was measured between Level 1-2, Level 2-3, Level 3-4, Level
4-5 (Figure 6), were monitored while the air pressure was applied inside the casing (Figure 36).
Case 1: Initial Condition (No pressure, Pi = 0)
The variation of initial resistance is shown in Figure 37. The initial resistance was higher
at the bottom (level 1-2) due to weight of the cement sheath and was lower at the top level (level
4-5). The electrical resistance increased with the depth and varied from 400 to 800 ohms. This is
partly due to the piezoresistive property of the smart cement, where the electrical resistance will
be higher with increase in pressure with depth.
35
Figure 37. Variation of initial resistance with depth after 100 days of curing
Figure 38. The changes in resistance with applied pressure for smart cement after 100 days
of curing
36
Case 2: Pi = 60 psi
Internal pressure of 60 psi was applied and the resistance changes were measured after 26
hours. The change in resistance was normalized with initial resistance R/R(%) is shown in
Figure 38. The resistivity change in the smart cement due to the applied pressure was about 0.5
to 0.6%, indicating the piezoresitivity of the smart cement.
Case 3: Pi = 100 psi
100 psi pressure was applied and the resistance changes were measured in 26 hours and reported
in the form of R/R (%) in Figure 38. The resistivity change in the smart cement due to the
applied pressure of 100 psi was about 2.5%, indicating the piezoresitivity of the smart cement.
Case 3: Pi = 140 psi
Internal pressure of 140 psi was applied and the resistance changes were measured in 26 hours
and reported in the form of R/R in Figure 38. The resistivity change in the smart cement due to
the applied pressure was about 6.5 to 7%, indicating the piezoresitivity of the smart cement.
Result and analysis
The stress at every point can be separated into mean stress and deviatoric stress. The
change in the deviatoric stress due to the applied pressure (Pi) along the axis of the casing (zaxis) is represented as Szz. Using equilibrium and stress analyses, it can be shown that Szz is
directly proportional to the applied internal pressure Pi. Hence the change in deviatoric stress can
be represeanted as follows:
Szz = f(Pi) ..(8)
The variation of internal applied pressure with the resistivity of smart cement (z) is
shown in Fig. 39, and the response of the smart cement is nonlinear.
p, q model
The nonlinear p-q model was developed by Vipulanandan et al. (1990) and was used to
predict z/z variation with the applied pressure. The relationship can be represented as follows:
(8)
37
The model parameters p and q were 1.2 and 3 respectfully. Hence measuring the change
resistivity of the smart cement it will be possible to predict the pressure in the casing and also the
stress in the cement sheath.
7
6
Expermintal Data
Model
/o (%)
5
4
3
2
1
0
0
20
40
60
80
100
120
140
160
Pressure (psi)
Figure 39. Model prediction of changes in resistivity with applied pressure for smart
cement after 100 days of curing
5. CONCLUSIONS
Based on the resistivity monitoring and model study following conclusions are advanced.
(1) Two probe method was effective in measuring the bulk resistance of the drilling mud, spacer
fluid and smart cement slurry. Based on the changes in resistance measurements it will be
possible to identify the fluid rise in the well borehole.
(2) Using the laboratory model it was possible to demonstrate the real-time monitoring of the
well bore with drilling mud, space fluid and smart cement slurry.
(3) Using the concept developed in this study, it is possible to use the K parameter and predict
the changes in the resistance of drilling and hardening smart cement.
(4) Using a nonlinear model the change in electrical resisitivity of smart cement was related to
the applied pressure in the casing. The smart ceent was very sensitive to the applied pressure.
38
6. BIBLIOGRAPHY
(1)
API Recommended Practice 10B (1997), Recommended Practice for Testing Well
Cements Exploration and Production Department, 22nd Edition.
(2)
API recommended Practice 65 (2002) Cementing Shallow Water Flow Zones in

Deepwater Wells.
(3)
Carter, K. M. and Oort, E. (2014), Improved Regulatory Oversight Using Real- Time
Data Monitoring Technologies in the Wake of Mocondo, SPE 170323, pp. 1-51.
(4)
Dom, P. B., S. Rabke, et al. (2007). "Development, verification, and improvement of a

sediment-toxicity test for regulatory compliance." SPE Drilling & Completion, Vol.
22(2), 90-97.
(5)
Griffith, J. and Faul, R. (1997) "Mud Management Special Slurries Improve Deepwater
Cementing Operations, Oil and Gas Journal, Vol. 95, No. 42, pp. 49-51.
(6)
Izon, D. and M. Mayes, M. (2007) Absence of fatalities in blowouts encouraging in

MMS study of OCS incidents 1992-2006, Well Control Magazine, pp. 86-90.
(7)
John B., (1992). Class G and H Basic Oil Well Cements, World Cement.
(8)
Kim, J and Vipulanandan, C. (2006)"Removal of Lead from Contaminated Water and

Clay Soil Using a Biosurfactant," Journal of Environmental Engineering, Vol. 132, No.
7, pp.857-865.
(9)
Kim, J. and Vipulanandan, C. (2003) "Effect of pH, Sulfate and Sodium on the EDTA
titration of Calcium," Cement and Concrete Research, Vol. 33(5), pp. 621-627.
(10) Kyle, M. and Van Eric (2014), Improved regulatory oversight using real- time data
monitoring technologies in the Wake of Mocondo, SPE 170323, pp. 1-51.
(11) Li, F., Vipulanandan, C. and Mohanty, K. (2003) "Microemulsion and Solution
Approaches to Nanoparticle Iron Production for Degradation of Trichloroethylene,"
Colloids and Surfaces A: Physicochemical and Engineering Aspects, Vol. 223, No. 1,
pp. 103-112.
(12) Labibzadeh, M., Zhhabizadeh, B. and Khajehdezfuly, A., (2010) "Early Age
Compressive Strength Assessment of Oil Well Class G Cement Due to Borehole
39
Pressure and Temperature Changes, Journal of American Science, Vol. 6, No.7, pp.3847.
(13) Nam, M. S. and Vipulanandan, C. (2008) "Roughness and Unit Side Resistances of
Drilled Shafts Socketed in Clay Shale and Limestone, Journal of Geotechnical and
Geoenvironmental Engineering, Vol. 134, No. 9 , pp. 1272-1279.
(14) Mirza, J. et al. (2002) "Basic Rheological and mechanical properties of High Volume
Fly Ash Grouts," Construction and Building Materials, Vol. 16, pp.353-363.
(15) Ozgurel, G., Gonzalez, H. A. and Vipulanandan, C., (2005) "Two dimensional Model
Study on Infiltration Control at a Lateral Pipe Joint using Acrylamide Grout",
Proceedings, Pipelines 2005, ASCE, pp. 631-642.
(16) Park, J., Vipulanandan, C., Kim, J. and Oh, M. H. (2006) "Effect of Surfactants and
Electrolyte Solutions on the Properties of Soils," Journal of Environmental Geology,
Vol. 49, pp.977-989.
(17) Ravi, K. et al. (2007) "Comparative Study of mechanical Properties of Density-reduced
Cement Compositions, SPE Drilling & Completion, Vol. 22, No. 2, pp. 119-126.
(18) Ramachandran, V. S. (1984) Concrete Admixture Handbook, Noyes Publication, Park
Ridge, New Jersey, 628 pp.
(19) Samsuri A., Junin R., Osman A.M. (2001). The utilization of Malaysian local bentonite
as an extender and free water controller in oil-well cement technology, Society of
Petroleum Engineers. Doi: 10.2118/68674-MS
(20) Thaemlitz, J., A. D. Patel, et al. (1999). "New environmentally safe high-temperature
water-based drilling-fluid system." SPE Drilling & Completion Vol. 14(3), 185-189.
(21) Vipulanandan, C. and Leung, M., (1991)" Seepage Control in Contaminated and
Permeable Houston Clay: Laboratory Study," Hazardous Waste & Hazardous
Materials, Vol. 8, No. 1, 17-32.
(22) Vipulanandan, C. and Neelam Kumar, M.(2000) "Properties of Fly Ash-Cement
Cellular Grouts for Sliplining and Backfilling Applications," Proceedings, Advances in
Grouting and Ground Modification, ASCE, GSP 104, Denver, CO, pp. 200-214.
(23) Vipulanandan, C., and Sett, K. (2004) "Development and Characterization of
Piezoresistive Smart Structural Materials", Proceedings, Engineering, Construction and
40
Operations in Challenging Environments, Earth & Space 2004, ASCE Aerospace

Division, League City, TX, pp. 656-663.
(24) Vipulanandan, C., and Liu, J. (2005) "Polyuretane Based Grouts for Deep Off-Shore
Pipe-in-Pipe Application", Proceedings, Pipelines 2005, ASCE, Houston, TX, pp. 216227.
(25) Vipulanandan, C., and Garas, V. (2006), "Piezoresistivity of Carbon Fiber Reinforced
Cement Mortar", Proceedings, Engineering, Construction and Operations in
Challenging Environments, Earth & Space 2006, Proceedings ASCE Aerospace
Division, League City, TX, CD-ROM.
(26) Vipulanandan, C. and Mamidi, B. (2008)Biosurfactant Flushing of PCE Contaminated
Clayey Soils, Proceedings, GeoCongress 2008, Geotechnics of Waste Management
and Remediation, ASCE, GSP 177, pp. 495-502.
(27) Vipulanandan, C. and Nam, E. (2009), Drilled Shaft Socketed in Uncemented Clay
Shale, Proceedings, Foundation Congress 2009, Contemporary Topics in Deep
Foundations, ASCE, GSP 185, pp. 151-158.
(28) Vipulanandan, C. and Usluogullari, O. (2009), Field Evaluation of A New Down-Hole
Penetrometer Proceedings, Contemporary Topics in In Situ Testing, Analysis, and
Reliability of Foundations, Foundation Congress 2009, ASCE, GSP 186, pp. 119-126.
(29) Vipulanandan, C., Dimrican, E. and Harendra, S. (2010) "Artificial Neural Network
and Nonlinear Models for Gelling and Maximum Curing Temperature Rise in Polymer
Grouts, Journal of Materials in Civil Engineering, Volume 23, No. 4, p. 1-6.
(30) Vipulanandan, C. and Prasanth, P., (2013)" Impedance Spectroscopy Characterization
of a piezoresistive Structural Polymer Composite Bulk Sensor," Journal of Testing and
Evaluation, Vol. 41, No.6, 898-904.
(31) Vipulanandan et al. (2014) Development and Characterization of Smart Cement for
Real Time Monitoring of Ultra-Deepwater Oil Well Cementing Applications, OTC25099-MS.
(32) Vipulanandan et al. (2014) Characterization of Smart Cement Modified with Sodium
Meta Silicate for Ultra-Deepwater Oil Well Cementing Applications, AADE-2014.
(33) Vipulanandan, C. Heidari, M., Qu, Q., Farzam, H., and Pappas, J. M. (2014),
Behaviour of piezoresistive smart cement contaminated with oil based drilling mud,
Offshore Technology Conference, OTC 25200-MS, pp. 1-14.
41
(34) C. Vipulanandan, C. and A. Mohammed, A. (2014), Hyperbolic rheological model

with shear stress limit for acrylamide polymer modified bentonite drilling muds,
Journal of Petroleum Science and Engineering, 122 (2014) 3847.
(35) Wang, S. Y. and Vipulanandan, C., (1996) "Leachability of Lead From Solidified
Cement-Fly Ash Binders," Cement and Concrete Research, Vol. 26, No. 6, pp. 895905.
(36) Zhang M., Sisomphon K., Ng T.S, and Sun D.J, (2010). Effect of superplasticizers on
workability retention and initial setting time of cement pastes, Construction and
Building Materials 24, 17001707.
(37) Zhang J., Weissinger E.A, Peethamparan S, and Scherer G.W., (2010). Early
hydration and setting of oil well cement, Cement and Concrete research, Vol. 40,
1023-1033.
42

Smart Cementing Drilling Muds Monitoring Deepwater Wellbore Enhancement

Hochgeladen von

Dokumentinformationen

Copyright

Verfügbare Formate

Dieses Dokument teilen

Dokument teilen oder einbetten

Freigabeoptionen

Stufen Sie dieses Dokument als nützlich ein?

Sind diese Inhalte unangemessen?

Copyright:

Verfügbare Formate

Smart Cementing Drilling Muds Monitoring Deepwater Wellbore Enhancement

Hochgeladen von

Copyright:

Verfügbare Formate

Smart Cementing Materials and Drilling

Muds for Real Time Monitoring of

Final Report Small Model Tests

Project No. 10121-4501-01

C. Vipulanandan and R. Krishnamoorti

Center for Innovative Grouting Materials and

Project 10121-4501-01 / UH-CIGMAT

Final Report-Small Model Tests

3. Theory and Concept................................................................................................... 7

Project 10121-4501-01 / UH-CIGMAT

Final Report-Small Model Tests

Final Report-Small Model Tests

Final Report-Small Model Tests

Project 10121-4501-01 / UH-CIGMAT

Final Report-Small Model Tests

Figure 1. Piezoresistivity Behavior of Oil Well Cement with 0.075% CF

1.2 Drilling Muds

Final Report-Small Model Tests

Final Report-Small Model Tests

A = effective cross sectional area

Project 10121-4501-01 / UH-CIGMAT

Final Report-Small Model Tests

Figure 3. Equivalent circuit for Case1

Figure 4. Equivalent circuit for Case 2

Project 10121-4501-01 / UH-CIGMAT

Final Report-Small Model Tests

Final Report-Small Model Tests

Figure 7: Different levels of cement in the oil well model

Final Report-Small Model Tests

Final Report-Small Model Tests

Figure 9. Horizontal resistance measurements with varying heights of water level

Project 10121-4501-01 / UH-CIGMAT

Final Report-Small Model Tests

Figure 10. Variation of resistance with rising water level

Final Report-Small Model Tests

Project 10121-4501-01 / UH-CIGMAT

Final Report-Small Model Tests

Figure 12: Resistance for free water and cement slurry

Figure 13. Horizontal slurry resistance with slurry level

Final Report-Small Model Tests

Project 10121-4501-01 / UH-CIGMAT

Final Report-Small Model Tests

Project 10121-4501-01 / UH-CIGMAT

Final Report-Small Model Tests

Project 10121-4501-01 / UH-CIGMAT

Final Report-Small Model Tests

Figure 17. Vertical resistance measurements for spacer fluid

Final Report-Small Model Tests

(c) Monitoring the Cement Slurry

Figure 18. Vertical resistance measurements for cement slurry

Project 10121-4501-01 / UH-CIGMAT

Final Report-Small Model Tests

Figure 19. Vertical resistance measurements in the oil well model

Project 10121-4501-01 / UH-CIGMAT

Final Report-Small Model Tests

Project 10121-4501-01 / UH-CIGMAT

Final Report-Small Model Tests

Figure 21: Variation of Parameter K for wire setup-a

Table 1: Variations of parameter K for wire setup-a

Std. dev. (m-1)

Figure 22: Variation of parameter K for wire setup-b

Final Report-Small Model Tests

Table 2: Variations of K parameter for wire setup-b

Std. dev. (m-1)