Vous êtes sur la page 1sur 18

SPE-196232-MS

Casing Rotation for Improved Cement Quality in Unconventional Horizontal


Wellbores

William Turner, Derek Adam, Kenneth Mike Cowan, Diego Tellez, and John Willis, Occidental Petroleum

Copyright 2019, Society of Petroleum Engineers

This paper was prepared for presentation at the SPE Annual Technical Conference and Exhibition held in Calgary, Alberta, Canada, 30 Sep - 2 October 2019.

This paper was selected for presentation by an SPE program committee following review of information contained in an abstract submitted by the author(s). Contents
of the paper have not been reviewed by the Society of Petroleum Engineers and are subject to correction by the author(s). The material does not necessarily reflect
any position of the Society of Petroleum Engineers, its officers, or members. Electronic reproduction, distribution, or storage of any part of this paper without the written
consent of the Society of Petroleum Engineers is prohibited. Permission to reproduce in print is restricted to an abstract of not more than 300 words; illustrations may
not be copied. The abstract must contain conspicuous acknowledgment of SPE copyright.

Abstract
Results from a long-term field study to improve zonal isolation show that rotating production casing in
unconventional well laterals is possible and that it delivers excellent cement quality. This paper reports the
results of the study. To date, 44 production strings have been rotated during cement jobs.
Specific details covered are cement bond evaluation results, methods used for modeling casing running,
torque and drag forces for floated and rotated strings, torque estimation for casing connection selection,
operational procedures, mud conversion cement additive, and best practices. Evaluation of cement bond
is based on ultrasonic cement logs acquired in casing strings cemented with and without casing rotation.
Advanced azimuthal ultrasonic logs were used to measure cement placement and quality. Trials with a mud
conversion cement additive evaluated whether the cement quality could be achieved with additives instead
of rotation. Both 6¾ in. and 8½ in. hole sizes in 10,000 ft lateral lengths were evaluated.
Prior studies demonstrated that casing movement is crucial for efficient mud removal (Hyatt et al. 1984;
Gai et al. 1996), yet it is not a common practice on long horizontal sections in our company. The casing
rotation initiative reported here was enabled by the availability of reliable high torque casing connections.
The results shared in this paper demonstrate that it is necessary to rotate casing while cementing horizontal
wells with 10,000 ft lateral lengths for effective cement isolation in the annulus. The cement bond resulting
from rotated cement jobs is near-perfect. Long-term study is needed to evaluate the effect of optimal cement
on hydraulic fracturing effectiveness, production, and the long-term health of the well.
Results reported include log data and modeled torque and drag values paired with operational data to
validate the models. Log data is available for eight lateral wellbores evaluated during the base case and trial
period. It is unique in that log data was acquired for both the base case and trial wells.
This paper supports well-known cementing best practices (McLean et al. 1967; Zurdo et al. 1986; Reiley
1987; Wilson et al. 1988; Sabins 1990; Torsvoll et al. 1991; Kettl 1993; McPherson 2000; Nelson et al.
2006; Al-Baiyat et al. 2019) with undisputable log data. The operational practices can be applied to many
unconventional operations to implement casing rotating into their cementing programs. The method shared
to estimate torque can be implemented by any drilling engineer with access to torque and drag software,
enabling them to choose the right casing connection for their application.
2 SPE-196232-MS

Introduction
In late 2017, a slim well architecture to improve well economics was evaluated. One fundamental
requirement of the slim well architecture was to retain 5½ in. production casing. Existing 10,000 ft lateral
well designs (referred onward as "10k laterals" in this paper) used 5½ in. casing in an 8½-in. borehole
(Willis et al. 2018). Determining the feasibility of both running 5½ in. casing in 6¾ in. hole and effectively
cementing it were the primary objectives of the slim design evaluation.
Preliminary hook load models indicated that floatation or rotation of casing would be required to get to
bottom. Casing standoff percentage would be much lower in the slim design than was achieved previously
in 8½ in. hole. There was concern how the new design would affect cement placement and isolation between
fracture stages.
To evaluate the new slim design and answer the questions related to running and cementing casing,
a thorough plan was developed and extensive engineering work was completed prior to drilling the first
slim trial well. The team acquired ultrasonic and cement bond logs (CBL) in both 8½ in. and 6¾ in. 10k
laterals, evaluated casing floatation technologies, and modeled hook load for floated casing runs. A method
to estimate torque while rotating during a cement job was developed using readily available software.
Knowledge of the operational torque range from the model assisted in casing connection selection. Lastly,
a repeatable process enabling regular rotation during cementing was developed and implemented. The log
results from the rotated cement jobs are presented showing the value of pipe movement, especially in low
standoff scenarios.

Establishing a Base Case


Prior to changing any aspects of current cementing practices, cement quality in current wellbores was
evaluated for both 6¾ in. and 8½ in. hole sizes. At the time, 10k lateral wells were 8½ in. hole size. Only
5,000 ft and 7,500 ft laterals had been drilled with 6¾ in. hole. Production cement slurries were designed to
adhere to industry best practices (McLean et al. 1967; Zurdo et al. 1986; Reiley 1987; Wilson et al. 1988;
Sabins 1990; Torsvoll et al. 1991; Kettl 1993; McPherson 2000; Nelson et al. 2006; Al-Baiyat et al. 2019)
with:

• Fluid loss < 70 mL/30 min

• Zero free water

• Transition time < 45 min

• Shear Bond > 300 psi

• Yield point < 15 lb/100ft2

• Cement rheology optimized for bottom hole temperature

In addition to the attention given to cement slurry design, centralizer placement and fluid flow regime
during placement were considered. Transitioning to 6¾ in. hole size reduced the casing standoff of the
existing one-per-joint centralizer placement scheme. The two plots below in Fig. 1 show the standoff
percentage for three different centralizer placement schemes: one-per-joint, two-per-joint, and three-per-
two-joints. At the time, 8½ in. laterals used one 8 in. outer diameter rigid body spiral centralizer per joint,
providing around 60% standoff. Using that same scheme with 6½ in. outer diameter centralizers in 6¾ in.
hole would reduce the casing standoff to 25%. In the small 6¾ in. × 5½ in. annulus, the difference between
25% and 60% standoff is 0.25 in. Installing more than one centralizer per joint shows clear improvement in
standoff percentage, but this benefit is offset by additional installation time, cost, and higher drag resulting
SPE-196232-MS 3

in longer casing running times. Weighing all these considerations, one centralizer per joint was the chosen
centralization scheme.

Figure 1—Casing standoff comparison for 6¾ in. and 8½ in. hole

Pipe movement and standoff percentage have a significant amount of interplay. Mud removal efficiency
decreases with standoff percentage as shown in Fig. 2. Over 70% standoff, pipe rotation is generally not
considered necessary. Between 45% and 65% standoff pipe rotation is advised, and below 45% rotation is
required for proper cement placement (Sanchez et al. 2012). No matter how high quality the cement slurry
pumped may be, achieving cement placement during the cement job and maintaining it afterwards is still
the most critical component of creating cement isolation.

Figure 2—Mud removal vs. standoff.


4 SPE-196232-MS

Predicting 20% standoff with one centralizer per joint in slim hole, falls squarely onto the red pipe
rotation required region in Fig. 2 above. Before committing to rotating casing during cement jobs, ultrasonic
and CBL logs were run to establish a reference quality for existing cement jobs completed with no pipe
movement. Logs were acquired over the full lateral length in wells of both hole sizes. The results are shown
in Figs. 3 and 4 below:

Figure 3—Base case ultrasonic cement log for 8½ in. hole.

Figure 4—Base case ultrasonic cement log for 6¾ in. hole.

Both logs reveal a low side channel with the channel in the 8½-in. wellbore appearing more continuous
than the channel in the 6¾-in. wellbore. Annular isolation in both cases varied along the wellbore. While
SPE-196232-MS 5

there were intervals with acceptable cement, a channel started to appear from mid lateral to the heel of the
wells. Both base case cement quality examples are not ideal; however, the 6¾-in. slim hole cement job
showed better overall annular isolation and lower CBL amplitude. Both log results demonstrate a need for
pipe rotation to eliminate the low side channel.
Seeing these channels extending across different intervals of the lateral it became apparent that poor
cement quality in horizontal wellbores could impact the quality of hydraulic fracture placement. While
plugs are used to isolate fracture stages inside the production casing, the primary role of cement in a
horizontal wellbore is to provide inter-stage annular isolation. With a channel present, a conduit exists for
subsequent fracture stages to re-enter previous stages resulting in uneven stimulation placement. When
this channel contains drilling fluid, the sand pumped could clean the channel replacing the mud with sand.
This uncemented conduit in the annulus may create a flow path along significant lengths of the wellbore
decreasing the efficiency of completions treatments. Long term, well productivity, ultimate recovery, and
well profitability could be adversely affected.
Of the two base case wells, only the 6¾-in. well had top of cement planned below surface. Fig. 5 shows
the slim hole log with a black dashed line indicating planned top of cement. Cement extends above the
planned top. The cement volume allocated for the low side of the casing has ended up above the planned
top of cement. With this realization, it is possible for anyone evaluating a horizontal CBL to recognize
significant cement above the planned top of cement as channeling and poor placement in the lateral.

Figure 5—Top of cement with no pipe movement.

Use of a Mud Conversion Cement Additive


Pipe movement is not the only solution to resolve poor low side non-aqueous mud removal. Another option
is a cement additive that converts non-aqueous fluid remaining after the cement job into a gelled sealant
(Kolchanov et al. 2018). The additive is incorporated into the dry cement blend and the job is pumped
conventionally. Being oil soluble, the additive reacts with the non-aqueous fluid it contacts and converts it
into a gelled sealant. Testing has shown it to be effective in various size channels, but it is most effective
in resolving smaller channels. In the paper referenced above, there were CBL examples showing improved
bond at the planned top of cement, but no evaluation was done in a horizontal wellbore.
6 SPE-196232-MS

To evaluate this new cement additive, it was added to the cement slurry for a horizontal production
casing and the well was cemented with no pipe movement and 20% standoff. The effect of the additive
on the anticipated low side channel was evaluated using an ultrasonic and CBL log. A common ultrasonic
log typically classifies material in the annulus in four ways: gas, fluid, cement, and microdebonding. Prior
to running the log, it was suspected that converted non-aqueous fluid would appear as microdebonding
with acoustic impedance higher than fluid but lower than cement. For any response to be the result of the
mud conversion cement additive, it should primarily appear in the region of the low side channel that was
identified on the base case logs.
The log response of the mud conversion cement additive appeared as predicted. Fig. 6 below shows a
portion of the log acquired and the low side channel near the heel is now visibly green with microdebonding
indicated. Cement quality from mid lateral to the toe of the lateral was very good, with almost no channel
present. From landing point to mid lateral, mud removal was less efficient resulting in the low side channel.
The additive was most effective over this region of the well that had the worst mud removal efficiency.
Frequently, cement logs show better cement quality at the casing joints, but the effect of the mud conversion
cement additive passes straight through the connections. This indicates the channel is continuous. The
variable density log (VDL) shows higher cement to formation bond quality with the additive when compared
with the base case shown in Fig. 5. The mud conversion cement additive improved the isolation between the
casing to cement and cement to formation interfaces and is a viable option for mitigating poor mud removal
efficiency resulting in a low side channel.

Figure 6—Cement log with mud conversion cement additive.

Slim Hole 10k Laterals and Floating Casing


Running casing to bottom consistently every time was a significant challenge in slimming down the well
design. Running 5½ in. casing in the 8½-in. 10k laterals was often challenging, requiring low friction
centralizers and occasionally fluid lubricant to land casing at planned depth. Evaluating the hookload
broomstick plot for running a 5½-in. casing in a 6¾-in. hole, as shown in Fig. 7, revealed that casing would
go to bottom only with a 0.2 friction factor (FF) and occasionally a 0.3 FF. At a 0.4 FF, the casing would not
make it to bottom without intervention. Expected hookload for 0.4 FF falls below the block and top drive
system weight at approximately 16,000 ft, which is 3,000 ft short of planned total depth (TD).
SPE-196232-MS 7

Figure 7—Conventional casing run hookload plot.

Rotating and or floating casing were considered as a secondary means to get casing to bottom consistently.
Rotating to bottom showed significant improvement on the broomstick plots; however, there was concern
about the effectiveness of rotation with free rotating centralizers every joint along the string. Rotating down
each stand would add 10 to 15 hours to each casing run. This additional time did not work with the well
economics so floatation was pursued.
Floatation is a straightforward process. The casing string is run in hole without filling to a predetermined
depth where a floatation sub is installed. Casing run above the floatation sub is filled with fluid while the
casing run progresses. Upon landing the casing, pressure is applied at surface to rupture the floatation sub.
Once ruptured, the fluid above the sub begins to fall and swap with the air in the lower portion of the casing.
Replacing the air-filled volume with fluid is done in multiple stages, allowing 10 to 15 minutes between
fluid swaps to allow air to evacuate. Pumping is not possible while floating, so the gel strength of the drilling
fluid can be broken using rotation every few thousand feet if necessary.
Torque and bending at the float sub require consideration to avoid unwanted casing failure. A torque
along string analysis is done for every run to ensure the floatation sub is positioned within its operational
limits. Float subs were run successfully above the curve and also a few hundred feet into the lateral. When
run in the lateral, the torque at the sub is lower. Avoid placing the floatation sub in any doglegs by evaluating
continuous inclination and azimuth data when determining its final position. To limit any unwanted bending,
position centralizers within 5 ft of the top and bottom of the float sub and avoid rotation while the float sub
is in build sections of the well.
Floatation was successful. Fig. 8 below shows the hookload broomstick plot for the floated casing run.
Actual hookload values, plotted as green circles, overlay a 0.2 to 0.3 FF. The torque and drag model
accurately predicted the hookload for the floating scenario. Hookload increased throughout the run and
there was no characteristic hookload decrease as casing was run into the lateral. The buoyant casing in the
lateral has less drag resulting from lower side forces, and the fluid weight added above the floatation sub
assists in running the casing.
8 SPE-196232-MS

Figure 8—Floated casing run hookload plot.

Estimating Torque While Cementing


Selecting a casing connection capable of rotation during the cement job was a significant challenge.
Engineering software was required to estimate the torque throughout the job, but common commercial
torque and drag software is not designed to estimate torque while fluids of different densities circulate the
wellbore in a dynamic condition. A method was developed to use common torque and drag software to
estimate the torque while rotating casing during a cement job. While there is software available to do this
at a more advanced level, this process should be available to all drilling engineers.
Torque loads change throughout the cement job. As cement fills the lateral section of the casing, side
forces increase resulting in a torque increase. Lighter displacement fluid replaces the heavy cement inside
the casing and the string once again becomes buoyant, lowering side forces and torque. If buoyancy was
the only factor, the point of highest torque during the cement job would be when cement reaches the shoe.
However, this is not the case. As spacer and cement begin to fill the annulus, the friction factor changes
as the fluid rheology changes. While pumping down the spacer and cement, the casing is rotating in oil-
based mud with a 0.28 friction factor (calibrated from actual data), but once spacer and cement enter the
annulus, 0.28 is no longer valid.
Torque during the cement job is estimated by evaluating surface torque at six different points throughout
the cement job. These points are:

• Circulating prior to cementing

• Cement at the shoe

• ¼ of cement volume in the annulus

• ½ of cement volume in the annulus

• ¾ of cement volume in the annulus

• Plug down – end of cement job


SPE-196232-MS 9

At each of these points, fluid positions in the casing and the annulus are entered in the software.
Casing and hole sizes have a significant effect of fluid position at these evaluation points. For example,
cementing a 5½-in. casing in an 8½-in. hole requires 470 bbl of cement to cover 500 ft above the previous
shoe. Cementing the same well with 6¾ in. hole requires 172 bbl of cement to cover the same interval,
approximately 300 bbl less than the 8½-in. hole. The fluid positions for these two cement jobs are very
different. In an 8½-in. hole, cement is exiting the shoe while it is still being blended and pumped. There is
a point when the entire production casing string is full of cement. In contrast, for a 6¾-in. hole, cement is
entirely downhole and the job is well into displacement before cement exits the shoe.
Identifying a common point in the cement job when maximum surface torque will occur is not possible.
The point of maximum torque depends on the casing, annular capacity, and planned cement volume. The
best way to assess the ability to rotate casing is to do the following torque analysis in the torque and drag
software.
Multiple open-hole wellbore sections are entered in the torque and drag software to account for different
friction factors for use with non-aqueous fluid (NAF), spacer, and cement filled wellbore sections. Friction
factors used are in Table 1 for the various fluids:

Table 1—Fluid specific open hole friction factor.

Hole Section Friction Factor Logic

Cased Hole 0.25 Calibrated to Actual Data

OH with NAF 0.30 Calibrated to Actual Data

OH with Spacer 0.30 to 0.40 sensitivity NAF FF to NAF FF +0.1

OH with Cement 0.30 to 0.50 sensitivity NAF FF to NAF FF +0.2

These values work well for the operating area evaluated by this paper; however, it is important for any
group doing this analysis to ensure the wellbore trajectory in the torque and drag software accurately reflects
true downhole tortuosity. Friction factor with NAF should be in the 0.25 to 0.30 range. Finding a friction
factor higher than 0.30 indicates the wellbore may be more tortuous than the survey indicates or suffers from
inadequate hole cleaning. If additional survey stations are not available to improve the well path description,
the higher friction factor carries through to the spacer and cement values as described in the logic column
of Table 1.
Fig. 9 shows example fluid positions, and the different open hole wellbore sections are indicated for each
of the six analysis points.
10 SPE-196232-MS

Figure 9—Fluid positions.


SPE-196232-MS 11

The following steps should be followed in the torque and drag software for each analysis point:

• Enter fluid positions and densities inside the casing

• Enter fluid positions and densities in the annulus

• Create separate wellbore sections for open hole intervals with drilling fluid, spacer, and cement

• Assign the correct friction factor or sensitivity to each wellbore interval

• Review the torque at surface plot and note down the minimum, average, and maximum estimated
torque.
At each fluid position, a torque at surface plot similar to the one shown in Fig. 10 below should be
generated. Fig. 10 resembles the analysis for rotating when one quarter of the cement is in the annulus. With
casing on bottom, only torque values at the total well depth are relevant yet the plot shows surface torques
along the entire well measured depth. Make note of the minimum, average, and maximum estimated torque.
Looking closely at the plot, it is possible to see the friction factor sensitivities for the different wellbore
intervals as they fan out.

Figure 10—Torque at surface.

After capturing minimum, average, and maximum estimated torque values at surface for all evaluation
points, a chart like Fig. 11 can be generated. Actual average surface torque values from the rotating cement
operation are plotted over the predicted values. One takeaway from these analyses is maximum torque never
occurs at the end of the cement job. If the string reaches maximum operational torque while pumping, keep
torque applied. As lighter displacement fluid enters the lateral, torque required will decrease and the string
occasionally starts rotating again as buoyancy increases.
12 SPE-196232-MS

Figure 11—Anticipated torque while cementing 5½ in. casing in 8½ in. hole

Wellbore Preparation Prior to Cementing


Wellbore preparation contributes to rotating during cementing success rate in addition to the overall quality
of the cement job. "When to start rotating?" is a question that came up many times during this initiative.
Should rotation checks be made while running casing to calibrate the torque model? Should casing be rotated
while circulating before the cement job? Lastly, should the casing be rotated while cement is being pumped
but has not exited the shoe? Most of these questions stem from discussion around potential risk of cyclical
fatigue failure introduced by rotating more than necessary.
Rotating at the different stages of cementing serves different purposes. Rotation while floating casing
in is important. In the procedure, rotation checks are made above the curve and every three thousand feet
throughout the lateral. No rotating should be done when the float sub is in the curve or any other localized
dogleg to reduce accumulation of fatigue at the floatation sub. These rotation checks acquire the torque
values needed to calibrate the model to see if rotation during cementing is possible. At this point, while still
floated, the torque is much lower. Most casing running procedures instruct to break circulation at a defined
interval, but circulation is not possible while floating casing. Instead, rotation breaks the fluid gel strength.
Lastly, the casing connection is made-up to optimal makeup torque, but continues to make-up down hole
as required. The resulting operational torque is higher than the makeup torque. While unusual, this is to
prevent torqueing connections to max operational torque unnecessarily. If the unfortunate situation arises
that casing be laid down, fewer joints are likely to be rejected and should be easier to break out. Rotating
every three thousand feet while running in hole incrementally torques the connections above makeup torque
as required ensuring easier rotation once on bottom.
Rotating during pre-cementing circulation continues to clean the wellbore and reduces the rotating torque
prior to starting the cement job. On the slim trial well, casing was rotated at 20 rpm while pumping 4 bpm
for ninety minutes. The rotating torques matched the calibrated torque at surface chart between a 0.2 and
0.3 FF. Interestingly, the torque shows a clear downward trend and then flattens out when plotted over time.
Fig. 12 shows the surface torque over the ninety minutes of wellbore cleanup. Maximum surface torque
reduced 10% from 18 kft-lbs to 16 kft-lbs. After a little over an hour, torque stabilized between 14 kft-lbs
and 16 kft-lbs indicating wellbore cleanup was complete. Field personnel did not see significant cuttings at
the shakers; however, rotation during this phase will remove fine solids and any gelled drilling fluid from
the low side of the well.
SPE-196232-MS 13

Figure 12—Wellbore conditioning torque reduction.

Casing is rotated at 20 rpm from the time the cement head is rigged up through the end of the cement
job. As cement travels down the casing, rotation and high fluid velocities continue to condition the annulus
prior to cement exiting the shoe. As discussed above, torque will increase as cement fills the casing in the
lateral portion of the well. As the torque increases, the connections will continue to makeup if casing is
rotating at this point. Starting to rotate as cement exits the shoe is more challenging. Since implementing
casing rotation, there has been a period of time during which rotation was not started until cement reached
the shoe. On these jobs, the rotation success rate during the remainder of cement placement was lower – 50
to 60% of the time vs. over 90% when casing was rotated prior to cementing and during the full cement job.

Rotation Rate
Higher rotation rates affect the leading cement to spacer interface more than cement placement. Base case
log results correlate closely with cement displacement simulations. Simulations at 5 rpm, 20 rpm, and 60
rpm evaluated the effect of rotation rate. Fig. 13 shows these results. The results among the three simulations
did not show any differences in cement placement. The primary effect of higher rotation rate was improved
cement to spacer interface. Increasing rotation rate from 5 rpm to 20 rpm showed considerable improvement.
Rotation at 60 rpm looked very similar to the results at 20 rpm. In Fig. 13 below, gray is cement, and the
two shades of green are spacer with slightly different properties. In the 5 rpm simulation, the two spacers
have completely intermixed, while they remain segregated in the 20 rpm and 60 rpm simulations.
14 SPE-196232-MS

Figure 13—Fluid segregation at different rates.

Rotating While Cementing – Log Results


Pipe rotation during wellbore preparation and throughout the cement job eliminated the low side channel.
The log results from the 8½-in. and 6¾-in. wellbores (see Figs. 14 and 15 respectively) with rotated and
centralized casing show near-perfect cement bond.

Figure 14—Rotated ultrasonic cement log in 6¾ in. hole.


SPE-196232-MS 15

Figure 15—Rotated ultrasonic cement log in 8½ in. hole.

Top of cement is also much closer to the anticipated top. Cement excess has been increased by 5% for
rotating cement jobs to ensure cement reaches the planned top. On the 8½-in. rotated casing trial, cement
did not meet the planned top of cement with the planned excess. The historical excess amount used was
sufficient for cement volumes that left a low side channel, but with cement fully placed in the annulus
additional cement was required. A top of cement comparison between the non-rotated and rotated slim hole
logs is shown in Fig. 16. As expected from the simulations shown in Fig. 13, the cement to spacer interface
is much tighter on the rotated cement job.
16 SPE-196232-MS

Figure 16—Slim hole top of cement comparison.

Rotating Casing: Other Considerations


Casing rotation complicates the cementing process and introduces risk not present during static cement jobs.
This risk must be evaluated and regardless of how clear the results are in the logs above, drilling groups and
companies that cannot bear the risk should consider other methods for cement bond improvement.
There have been some serious lessons learned regarding centralizers that need to be passed on. First,
polymer centralizers also referred to as low friction centralizers should not be used if casing is to be rotated.
Considerable amounts of polymer material returned to surface confirming that they do wear down while
running in. In slim hole, the spiral wrap is insufficient to keep the body of the centralizer from riding on
the low side of the hole. It is possible for these to wear through (Rodrigue et al. 2019). Once rotation is
started, the centralizers can deform, pack-off and restrict flow in the annulus. Second, a stop collar must
be run above and below any centralizer when running semi-flush casing connections, even with steel body
centralizers. The leading and trailing edge of the centralizer must meet a flush surface like the edge of a
casing coupling or stop collar. When forcing up centralizers against the upset of a semi-flush connection,
they expand, thus increasing drag and obstructing the annulus. Lastly, centralizers aid in rotation. Cement
jobs rotated without centralizers consistently rotate at higher torque than centralized strings.
Production casing strings can contain many accessories that need to be evaluated for integrity during
rotation. During rotation, these accessories experience tensile and compressive loads, fluctuating torsion,
and cyclic bending loads. It is recommended to have third party finite element analysis performed on
any casing accessories that experience loads near their ratings. Every well trajectory is different from a
design and an as-drilled standpoint. Torque along the string modeling should be done to ensure components
are positioned within their operating limits. Continuous inclination and azimuth data has been helpful in
avoiding localized short course length doglegs when placing the floatation subs.
This paper focuses on rotating casings in horizontal wellbores, but there is equal if not more value in
rotating casing in vertical wells. Rotation in vertical well sections should provide lasting benefits to the well
for life. Cement should be placed as desired, and serve its intended purpose: preventing fluid migration to
shallower formations, protecting water zones, protecting the casing from corrosive fluids, and preventing
SPE-196232-MS 17

migration of injected fluids in enhanced oil recovery projects. Rotation in vertical well sections can also be
accomplished at considerably lower surface torque and cost.

Conclusion
1. Incorporating casing rotation into the slim well design was justifiable from both an economic and
wellbore quality standpoint. The results in Figs. 14 and 15 confirm that rotating casing during
horizontal cementing provides a step change improvement in cement quality.
2. Slimming down the well architecture lowered overall well cost by approximately 18% and production
casing floatation and rotation made it possible to deliver the same production casing size with better
cement quality than the preceding big hole wells. In this implementation, capital reduction offsets the
incremental cost to rotate casing, which is less than 1% of planned well cost.
3. Cement quality is near-perfect on ultrasonic cement evaluation logs when 5½ in. casing is rotated in
laterals on both 8½ in. and 6¾ in. hole sizes.
4. Casing rotation while circulating prior to cementing can remove additional cuttings from the well.
5. Cementing quality with oil-based mud can be improved by a mud conversion cement additive that
reacts with the oil in the mud.
6. Rotating casing on 10k laterals requires careful analysis of torque and drag.
7. Casing connections must be capable of the torque requirements for rotation.
8. Centralizers reduce rotating torque and improve cement quality when rotating.
9. Polymer centralizers should not be used when rotating casing due to wear.
10. Floating casing in 10k laterals reduces drag significantly and increases the speed of running casing
and the certainty of reach total depth.
11. Wells with rotated cement jobs should be tracked closely to enable future evaluation whether the
improved isolation improves hydraulic fracture placement and stimulation, affects initial production
and ultimate recovery, and improve re-fracturing.
The value of optimal cement quality likely extends far beyond this capital expenditure reduction into
lowering operating expense for the life of the well. To date, 44 laterals 10k or longer have successfully
rotated cement jobs and over 60 wells used the mud conversion cement additive. That is nearly half a million
feet of producing reservoir with optimal cement quality.

Al-Baiyat, I., Khalil, M., Heinhold, T. et al 2019. Production Liner Cementing Practices in Challenging Horizontal
Wellbores. International Petroleum Technology Conference, Beijing, China, 26-28 March. DOI: 10.2523/IPTC-19471-
MS.
Gai, H., Summers, T. S., Cocking, D. A. et al 1996. Zonal Isolation and Evaluation for Cemented Horizonal Liners. SPE
Drilling and Completion, December. DOI: 10.2118/29981-PA.
Hyatt, C. R. and Partin Jr, M. H. 1984. Liner Rotation and Proper Planning Improve Primary Cementing Success. SPE
Deep Drilling and Production Conference, Amarillo, Texas, USA, 1-3 April. DOI: 10.2118/12607-MS.
Kettl, F. C., Edward, M. G., and Covington, R. L. 1993. Practical Horizontal Cementing Today. Middle East Oil Show,
3-6 April, Bahrain. DOI: 10.2118/25546-MS.
Kolchanov, P., Perroni, D., Medvedev, A. et al 2018. Effective Zonal Isolation in Horizontal Wells: Mitigating Negative
Impact of Mud Channels. SPE Annual Technical Conference and Exhibition, 24-26 September, Dallas, Texas, USA.
DOI: 10.2118/191561-MS.
McLean, R. H., Manry, C. W., and Whitaker, W. W. 1967. Displacement Mechanics in Primary Cementing. Journal of
Petroleum Technology, Volume 19, Issue 02, February. DOI: 10.2118/1488-PA.
McPherson, S. A. 2000. Cementation of Horizontal Wellbores. SPE Annual Technical Conference and Exhibition, 1-4
October, Dallas, Texas. DOI: 10.2118/62893-MS.
Nelson, E. and Guillot, D. 2006. Well Cementing. Second Edition. Sugar Land, Texas: Schlumberger.
Reiley, R. H. 1987. Cementing of Liners in Horizontal and High-Angle Wells at Prudhoe Bay, Alaska. SPE Annual
Technical Conference and Exhibition, 27-30 September, Dallas, Texas. DOI: 10.2118/16682-MS.
18 SPE-196232-MS

Rodrigue, M., Kendziora, L., and Farley, D. 2019. Myth-Busting Performance Properties of Nonmetallic Rigid
Centralizers. SPE/IADC International Drilling Conference and Exhibition, 5-7 March, The Hague, The Netherlands.
DOI: 10.2118/194094-MS.
Sabins, F. 1990. Problems in Cementing Horizontal Wells. Journal of Petroleum Technology, Volume 42, Issue 04, April.
DOI: 10.2118/20005-PA.
Sanchez, A., Brown, C., and Adams, W. 2012. Casing Centralization in Horizontal and Extended Reach Wells.
SPE/EAGE European Unconventional Resources Conference and Exhibition, 20-22 March, Vienna, Austria. DOI:
10.2118/150317-MS.
Torsvoll, A., Olaussen, S. R., and Almond, S. W. 1991. Keys to Successful Cement Job for a Horizontal Liner on Statfjord
A Platform, Well A-37A: A Case Study. SPE/IADC Drilling Conference, 11-14 March, Amsterdam, Netherlands. DOI:
10.2118/21970-MS.
Willis, J., Tellez, D., Neel, R. et al 2018. Unconventional Drilling in the New Mexico Delaware Basin Case History.
IADC/SPE Drilling Conference and Exhibition, 6-8 March, Fort Worth, Texas, USA. DOI: 10.2118/189597-MS.
Wilson, M. A. and Sabins, F. L. 1988. A Laboratory Investigation of Cementing Horizontal Wells. SPE Drilling
Engineering, Volume 3, Issue 03, September. DOI: 10.2118/16928-PA.
Zurdo, C., Georges, C., and Martin, M. 1986. Mud and Cement for Horizontal Wells. SPE Annual Technical Conference
and Exhibition, 5-8 October, New Orleans, Louisiana. DOI: 10.2118/15464-MS.

Vous aimerez peut-être aussi