Sie sind auf Seite 1von 7

CTOver.

qxd

7/11/03

8:52 AM

Page 50

OVERVIEW
Cementing Technology
Not so long ago, the mantra of virtually every cementing operation was
pump it, bump it, and go to the house. The only design emphasis was
being sure there was enough thickening time to get the cement in place,
possibly include some fluid-loss control, and have it set in some reasonable period of time.
Most cement studies related to how to improve on the placement process
of the cement to make it more efficient. Researchers continue to look at
mud displacement, pipe movement, centralization, and a host of techniques aimed at
improving the proper placement of the cement-critical areas that remain important today.
Benge

Gas migration came to the forefront as a major concern. Companies researched the problem and developed a myriad of solutions. When the problem appeared to be resolved,
drilling in the deep waters of the Gulf of Mexico showed us the challenges of cementing
in an environment with shallow water flows. Foamed cement was found very effective in
the control of shallow water flows, and the use of foamed cement has become routine in
offshore applications, a technology many thought too complex for that environment a few
years earlier.
The realization now is for the need of cement designs that will maintain the seal in the
wellbore throughout the life of the well. Trying to develop models to predict the stresses
in the well throughout the full productive life is a daunting task, but one that will
be resolved.

Additional
Cementing
Technology
Technical Papers
Available at the SPE e-Library:

www.spe.org
SPE 79907
Advances in Tieback
Cementing
SPE 77753
Subsea Remedial Cementing
of Conductors on TemplateDrilled Development Wells
SPE 81182
Challenging the Limits: Setting
Long Cement Plugs

Sophisticated wellbore-stress analysis is being performed, and the output from these simulations is used to custom design the cement slurry. Cement designs are custom fit to individual wellbore-stress environments that change during the lifetime of the well.
Research into the long-term rock mechanics properties of cements is revealing new
insights into technology required to maintain a lifetime seal in the wellbore. The learning
curve is quite steep, and as more is learned, more questions are raised.
We have grown beyond the simplistic pump, bump, and go home mode. Placing cement
in the well is only the beginning.
JPT
Glen Benge, SPE, is Engineering Associate with ExxonMobil Development Co. in Drilling
Technical Applications. He oversees the cementing technology for ExxonMobils worldwide
drilling operations. He has 26 years experience associated with all aspects of cementing
and has authored many papers on cementing design and application. Benge has served on
the SPE Forum Series North America, Annual Technical Conference, and Drilling
Conference Program Committees and is on the JPT Editorial Committee.

50

AUGUST 2003

77754.qxd

7/11/03

7:45 AM

Page 51

C e m e n t i n g Te c h n o l o g y

A Model of Hydrostatic Pressure Loss


in Wells After Cement Placement
Hydrostatic pressure loss after cement
placement is considered to be a cause of
early gas migration. In high-pressure
zones, hydrostatic pressure loss can lead
to gas invasion and flow in the cemented
annulus. The pressure-unloading model
developed in the full-length paper was
validated using data from published field
experiments. The theoretical results
matched experimental data showing an
exponential trend in the downhole pressure loss.
Introduction
Several accidents have occurred on Outer
Continental Shelf wells shortly after cementing while rig personnel waited for the cement
to set. Data from these incidents have been
accumulated by the U.S. Mineral
Management Service during the last 20 years.
In all these events, the diverter system was not
completely nippled down before the cement
developed sufficient compressive strength.
Typically, a few hours after cement placement,
a well would start flowing gas. This type of
flow is difficult to control and, if not eliminated, may cause loss of well integrity, spillage, or
a blowout. The problem is known as early gas
migration or shallow gas flow.
Because of limited data, analysis of field
cases of flow after cementing has been limited to qualitative observations about common features of the affected wells. Although
not unique to wells that experience gas flow,
some common features include:
Casing strings are set at shallow depths.
Hole sizes are relatively large.
Cement slurries are displaced to the
surface.
Average time to flow was 4.6 hours after
cement placement. Several wells experienced the problem at the same time that
cement slurry displacement was complete.
Depths of these wells averaged 3,412 ft,
while occurrence in wells deeper than
5,000 ft was rare. Casing sizes ranged
from 103/4 to 12 in.
A typical sequence of events leading to
flow after cementing can be summarized
by the following events.
1. Cement is pumped and displaced
successfully.

AUGUST 2003

2. After a few hours of waiting on


cement, the diverter/blowout preventer
stack is nippled down and the well starts
to flow.
3. The diverter is nippled up again and
the well is diverted in an attempt to control the flow.
4. The well is still flowing and becomes
more difficult to control.
5. Attempts are made to gain control of
the well.
6. In cases of severe flow, the rig is
evacuated.
Incidents of flow after cementing are
caused by the presence of shallow gas
pockets tapped by the wells. These pockets may go undetected during drilling
because their pressure gradient is less
than the drilling fluid equivalent pressure
gradient. Mechanistic analysis and mathematical modeling are necessary to understand and describe gas flow observed after
cementing. Several concepts have been
used to describe hydrostatic pressure loss
in cement during the dormant and transition stages of slurry setting. All use a static approach with time-dependent friction
based on static gel strength development
over time. However, there is a lack of consistent theory that considers the timedependent motion of the slurry and the
associated frictional pressure-loss effects
in the pressure-loss phenomenon.
Field Measurements
One investigator published downhole pressure measurements taken to investigate
causes of gas migration behind the casing
after primary cementing in full-scale wells.
Six sensors were attached to the casing at
various depths to record pressure and temperature. The hole size was 77/8 in., and
27/8-in. casing was run to 8,900 ft. Nearly
400 bbl of 16.6-lbm/gal cement slurry was
pumped. Drilling mud density was
10.2 lbm/gal. Hydrostatic pressure
dropped exponentially in early time after
cementing. An almost 3,000-psi pressure
drop was observed at 8,754 ft. Pressure
sensor recordings indicated that the
hydrostatic pressure in the annulus converged to the formation pressure gradient.

This article, written by Assistant


Technology Editor Karen Bybee, contains highlights of paper SPE 77754,
Transient Pressure UnloadingA
Model of Hydrostatic Pressure Loss in
Wells After Cement Placement, by
Somei Nishikawa, SPE, Mitsui Oil
Exploration Co., Ltd., and Andrew K.
Wojtanowicz, SPE, Louisiana State U.,
prepared for the 2002 SPE Annual
Technical Conference and Exhibition,
San Antonio, Texas, 29 September2
October.
For a limited time, the full-length paper
is available free to SPE members at
www.spe.org/jpt.

During slurry setting, the cement level


dropped 300 ft. The temperature log
showed the top of cement at 1,500 ft below
the rotary table, while the calculated depth
of the top of cement was 1,200 ft.
Hydrostatic pressure loss resulting from the
300-ft cement height change is calculated to
be 259 psi, yet the bottom sensor indicated
a 2,950-psi pressure loss. Less than 10% of
the measured pressure loss could be attributed to fluid level change.
Objectives
One objective was to examine hydrostatic
pressure loss as a result of a propagating
pressure transient caused by a decrease in
cement slurry volume downhole. The volume reduction mechanism was analyzed
and included in a mathematical model of
bottomhole pressure change with time.
Another objective was to determine if the
duration of the pressure transient effect is
sufficiently long to match field observations
of gas flow at the surface. Another objective
was to match the predicted pressures with
recorded downhole pressure changes to validate the model.
Pressure Transient Unloading Model
The pressure unloading model considers
the pressure transient effect resulting from
propagation of the volumetric reduction

51

77754.qxd

7/11/03

7:46 AM

Page 52

recorded. The conclusion was


caused by bottomhole fluid
reached that chemical shrinkloss. After the cementing job
age occurred earlier before stais completed, the cement slurtic gel strength increased. Conry column in the openhole
sequently, chemical shrinkage
annulus loses water to the
did not contribute to hydrostarock because of pressure overtic pressure loss.
balance and filtration. As the
fluid loss continues, part of
Static Gel Strength Assumpthe cement slurry moves
tion. The model assumes condownward in plug flow. This
stant rheology of the cement
downward movement generslurry. After experimental
ates friction at the annular
studies of static gel strength,
walls that reduces the hydrothe model was modified to
static pressure. A pressure
ignore any effect of static gel
unloading transient effect is
Fig. 1Measured and calculated hydrostatic pressure loss at
strength development on
generated and transmitted
depth.
hydrostatic pressure. At early
from the fluid-loss zone to the
time, the static gel strength
surface. As a result, the preseffect is much smaller than the
sure decrease along the
effect of fluid loss. At later
cement column is a function
time, the static gel strength
of time and depth. The foleffect becomes large and
lowing assumptions were
should affect the hydrostatic
made in deriving the mathepressure loss.
matical model.
There is only one fluidModel Verification
loss zone and one gas zone in
The model was validated by
an open hole, and the fluiduse of published annular presloss zone is near the gas zone.
sure and temperature data
Only the fluid loss confrom field tests. Fig. 1 shows a
tributes to volumetric change.
comparison of measured
At early time, chemical
downhole pressure losses and
shrinkage is small and does
Fig. 2Hydrostatic pressure distribution with depth.
pressure losses computed with
not cause a hydrostatic presthe model. The model provides
sure loss.
The hydrostatic pressure loss continues cement setting time, the rod weight was an excellent match with errors less than 10%.
Fig. 2 shows the downhole pressure disuntil it balances the pressure of the fluid- measured constantly. Chemical shrinkage
loss zone.
should cause downward movement of the tribution at different times. The plots are
The fluid-loss rate is constant and con- cement slurry, while gelation should almost linear with a slight curve at the bottrolled by the effect of decreasing overbal- increase friction at the rod surface. If the tom representing the tail slurry. The match
ance on cake permeability.
two phenomena occur concurrently, the rod with recorded pressure is excellent.
The model also was verified by compar Cement slurry density and compress- should be pulled down by a friction force
ibility are constant.
that would add to the measured weight. ing the effects of fluid loss and pressure
A plug-flow model describes cement Although the experiment design did not unloading. Fluid loss was calculated using
slurry motion.
simulate the entire hydrostatic pressure loss an average hydrostatic pressure change.
Cement slurry rheology is constant dur- process in an actual well, it simulated the The 270-ft calculated vertical cement
ing the early time after cement placement, possible relationship between chemical height change agrees well with the 300-ft
and gel strength is constant during the dor- shrinkage and pressure transmission in the cement level drop measured in the field
experiment. Hydrostatic pressure loss at the
mant period of slurry setting.
well annulus.
The gas-zone pressure is higher than
The first experiment used a 12-in.-long cement bottom results from fluid loss and
the fluid-loss zone pressure.
solid rod placed in the cylinder filled with system compressibility.
An appendix in the full-length paper cement slurry. The wetted length between
shows the derivation of the equation for the the solid rod and the cement slurry was 6 in. Conclusions
1. Chemical shrinkage does not control
pressure unloading model and the fluid- In the second experiment, a 36-in.-long solid
loss volume.
rod was placed in the cylinder. Wetted length hydrostatic pressure loss after cement placein the second experiment was 31 in. After ment.
2. The new pressure unloading matheExperimental
380 min, the average static gel strength was
An apparatus was built to investigate the 500 lbf/100 ft2. Rod weight was estimated to matical model was validated by use of pubfriction force caused by chemical shrinkage increase by 61 g after 380 min. The cement lished field test data.
3. The match of pressure and downhole
and static gel strength development. A 1/4-in. top was observed to drop by 1/4 in. and 3/4 in.
solid steel rod was placed in a 23/4-in. inside in the first and second experiments, respec- fluid loss with data measured in actual wells
diameter cylindrical container filled with tively, indicating that chemical shrinkage suggests that the model represents down15.9-lbm/gal Class H cement. During occurred. However, no weight change was hole fluid loss accurately.
JPT

52

AUGUST 2003

77804.qxd

7/11/03

7:48 AM

Page 53

C e m e n t i n g Te c h n o l o g y

Innovative Through-Tubing Cement


Packer Technique
Cement packers have been used for
many years to access behind-pipe
reserves above existing completions in
wells with multiple completion targets.
The basic function of the cement packer
is to create a new annular barrier
between an existing production packer
and the surface. Many cement packer
operations are performed with coiled
tubing (CT) or hydraulic workover
(HWO) units. The full-length paper presents a technique that can be used to
perform a cement packer operation
using only wireline (slickline and electric
line) and pumping services.
Introduction
Use of a single wellbore to access multiple
completion targets is a cost-effective way to
maximize recovery while minimizing cost.
In most cases, target reservoirs are accessed
from bottom to top. The initial completion
is designed to allow access to the upper
zones of interest without requiring the tubing to be pulled. In some circumstances,
the initial completion may require the completion packer to be set below the next zone
of interest.
The objective of a cement packer recompletion is to place a cement plug in the
annulus above the existing mechanical
packer. In many applications, CT or a HWO
unit is used to place the cement packer. In
some locations, CT or HWO equipment
cannot be used because of physical constraints (platform size, crane capacities,
equipment availability) or poor economics.
The cement packer technique discussed is a
process that can be used to perform a
cement packer recompletion using only
wireline (slickline and electric line) and
pumping services. By eliminating the use of
CT or a HWO unit, the overall cost and
complexity of the project can be reduced.
Screening
The first step in selecting a candidate for
this cement packer recompletion technique
is proper screening. Without proper screening, problems can occur during the cement
packer operation that can negatively affect
well performance.

AUGUST 2003

Wellbore Configuration. The wellbore


should allow abandonment of the existing
completion without interfering with other
completed or target zones. Typically, candidate wells are single completions. Another
requirement is that circulation must be possible up the annulus between the production tubing and casing. Reasonable cement
bond should exist behind the casing on
either side of the zone of interest.
Future Wellbore Use. Before performing a
cement packer recompletion, a review of the
future use of the wellbore must be completed. In most cases, this requires knowledge
of the depletion plan for an area or field. No
future work requiring removal of the tubing
below the cement packer location should
be planned.
Well Status. Ideal candidate wells are shutin wells, wells that produce at or below their
economic limit, or wells that produce at a
very high water cut. Recent completions
and idle wells in which the tubulars have
been properly protected also are good candidates. In addition, the wellbore should
have tubing integrity above the existing permanent packer.
Operational Process
Preliminary Work. To verify wellbore conditions, slickline is used to check the tubing
inside diameter (ID) and prepare the well to
be pressure tested. Gauge runs of various
sizes are made. Gas lift valves or orifices
installed in the tubing string should be
replaced with dummy valves. A retrievable
plug is set deep in the tubing (typically
below the existing packer, but above the
existing completion). The tubing is tested
against the plug to confirm tubing integrity.
A check also should be made to verify that
all surface control valves are functioning
properly. The final step in the preliminary
work process is to run a static-temperature/-pressure survey. This information will
be used to help determine the cement slurry specifications.
Abandoning an Existing Completion.
Electric line is used to set a mechanical plug

This article, written by Assistant


Technology Editor Karen Bybee, contains highlights of paper SPE 77804,
Innovative Through-Tubing Cement
Packer Technique, by M.L. Harris,
SPE, Dubai Petroleum Co., prepared for
the 2002 SPE Asia Pacific Oil and Gas
Conference and Exhibition, Melbourne,
Australia, 810 October.
For a limited time, the full-length paper
is available free to SPE members at
www.spe.org/jpt. The paper has not
been peer reviewed.

above the existing completion; cement is


then placed on top of the plug using a dump
bailer. Depending on the location of the targeted interval, the mechanical plug can be
set in the casing below the tubing or in the
tubing. Two types of mechanical plugs are
used to set in the casing below the tubing:
an inflatable-type plug and a compression-type plug. Both plugs are sufficiently
small to pass through the tubing and
expand to the ID of the casing.
Several options are available for plugs set
in the tubing. One option is to run a tubing
bridge plug and set it at the desired location.
A second option is to set a plug in a deep
nipple profile above the existing completion
and below the zone of interest. Cement
should be placed on top of the plug to form
a secondary barrier and improve the pressure differential across the plug.
The cement slurry should be designed to
allow sufficient time to mix and run to the
desired dumping depth. Only the cement
required for one bailer run should be mixed
at a time. To simplify the mixing process
and maintain consistency between the
cement slurries for each bailer run, the
cement and additives should be premeasured into kits. Each kit should consist of
the volume of cement and additives necessary for one bailer run.
The cement is placed on top of the plug
using a dump bailer. Once the cement is set,
the plug is pressure tested to the maximum
anticipated surface pressure.

53

77804.qxd

7/11/03

7:48 AM

Page 54

Tubing Punches. After abandoning the


existing perforations, the tubing is perforated using a tubing puncher gun. The puncher gun will penetrate the tubing but not the
casing. The tubing punches will be used to
circulate the cement packer into the annulus. Ideally, the tubing punches should be
located below the recompletion zone.
Cement Packer Placement. Before pumping the cement packer, an acid tubing wash
should be performed to eliminate any contaminants on the tubing wall that may cause
the cement to set up prematurely. The tubing wash should be circulated down the tubing to the tubing punches while taking
returns up the tubing/casing annulus.
Normally, an acid soak time also is included
as a part of the tubing wash to help remove
as many contaminants as possible.
The cement packer volume should provide a minimum of 400 ft of cement packer
thickness in the tubing/casing annulus. This
thickness is based on a 100-ft plug with 50 ft
of contaminated cement on both sides using
a safety factor of 2. All the cement packers
pumped using this technique, with a thickness within this range, have been successfully tested to at least a 2,500-psi differential.
The cement slurry should be placed a
minimum of 100 ft above the tubing punches in the tubing/casing annulus to account
for any discrepancies in volume during
pumping the cement packer. The maximum
observed discrepancy in cement packer
placement has been approximately 50 ft.
Before pumping the cement slurry, the
well should be circulated with 1 to 2 tubing
and tubing/casing annulus volumes. The
wellbore circulation can be performed while
batch mixing the cement if all the necessary
pumping equipment is available. A tubingmovement calculation should be performed
to analyze the tubing stress for all future
wellbore conditions.
Immediately before pumping the cement
slurry, all the wing and header valves should
be closed. The swab valve, master valves,
and return line from the tubing/casing
annulus should be opened. The cement
slurry should be displaced at a rate sufficient to maintain turbulent flow in both the
tubing and tubing/casing annulus. To minimize contamination, the cement slurry
should have a 10-bbl fresh-water spacer
before and after the cement slurry. Once the
cement slurry reaches the tubing punches a
noticeable increase in treating pressure will
be observed. When all the cement has exited the tubing punches, a drop in the treating pressure will be observed. When this
pressure drop occurs, the tubing/casing

54

annular volume necessary to place the bottom of the cement packer at the desired
location in the tubing/casing annulus
should be pumped.
Immediately after placing the slurry at the
desired location, all valves on the wellhead
should be closed. Closing the valves prevents movement of the cement packer.
Once the cement has set up, slickline gauge
runs should be made to verify there are no
restrictions in the tubing. A temperature
survey also should be run to verify the location and quality of the cement packer.
Finally, the cement packer should be
pressure tested. Test pressure should be
equivalent to the maximum expected differential from gas lift, stimulation, or production conditions.
Case Studys
Over a 5-year period, several recompletion
opportunities were identified in a Middle
East reservoir with marginal reserves. These
recompletion opportunities were not pursued because of the marginal economics
associated with bringing a rig on site to perform the work.
Phase IPilot Program. In late 1999, a
pilot plan was developed to perform two
cement packer recompletions using only
wireline (slickline and electric line) and
pumping services. Both of the proposed
wells were located on the same platform and
were already shut in.
Well 1. The target recompletion zone in
Well 1 was completely above the existing
packer. The plan for Well 1 was to abandon
the existing perforations by performing a
bullhead cement squeeze down the tubing
before placing the cement packer.
After executing the bullhead cement
squeeze, cement height in the tubing was
found 790 ft above the desired height. The
improper cement height was attributed to
either contaminants on the tubing wall,
improper metering of the displacement volumes, or backflow from the reservoir after
the cement squeeze.
The platform size was sufficient to allow
CT to be rigged up after making equipment
modifications to the unit. After rigging up
CT, the cement in the tubing was drilled out
to a sufficient depth to allow the recompletion process to continue without any additional need for CT.
The tubing was punched with 2 shots/ft
(SPF) across a 5-ft interval below the zone of
interest. Cement slurry was pumped down
the tubing and displaced at a rate of 8 bbl/min
to generate a 700-ft cement packer in
the 41/295/8-tubing/casing annulus. After

allowing the cement to cure, the cement


packer was successfully tested to 2,500 psi.
The target zone was perforated and acid stimulated. The dummy valves were pulled from
the gas lift mandrels and replaced with gas lift
valves. Finally, the well was placed on production and tested at economical rates.
Well 2. In Well 2, the target zone was
above and below the existing permanent
packer. The abandonment procedure for
Well 2 was to set an inflatable bridge plug
and dump bail cement on top of the plug
using electric line before pumping the
cement packer.
An inflatable plug was set successfully in
Well 2, and a 10-ft cement plug was dump
bailed on top. The tubing was punched
across 5 ft at 2 SPF. Because of the location of
the existing packer, the tubing punches were
located in the middle of the proposed recompletion target interval. A 700-ft cement packer was successfully placed into the
tubing/casing annulus by circulating the
cement slurry down the tubing. While testing the cement packer to 2,500 psi, the
cement plug differential was exceeded. The
plug failed, and communication was reestablished with the abandoned perforations.
A compression-type plug was run on
electric line in place of the inflatable plug to
prevent movement of the plug downhole.
While running in the hole, the plug became
stuck in the tubing but was successfully
fished out with slickline. An inflatable-type
plug was chosen for the third attempt.
Using electric line, an inflatable bridge plug
was run and appeared to set successfully in
the well. Because the CT equipment was
still on the platform, it was used to spot a
50-ft cement plug on top of the inflatable
bridge plug. Sometime during the CT operation, the inflatable bridge plug apparently
failed, and no cement height was established. The CT equipment was rigged
down, a fourth inflatable bridge plug was
set, and 20 ft of cement was successfully
dump bailed on top of the plug. A successful pressure test to 3,000 psi was performed.
The recompletion interval was perforated
and stimulated. While performing the stimulation treatment, the breakdown pressure
reached approximately 4,000 psi. Throughout the treatment, no indication of communication past the cement packer was observed.
After completing the stimulation, the
dummy valves were retrieved from the gas
lift mandrels and replaced with gas lift
valves. The well was placed on production
and tested at economic rates. Although complications had developed requiring the use of
CT in both wells, the CT was not used in the
JPT
placement of the cement packer.

AUGUST 2003

77756.qxd

7/11/03

7:46 AM

Page 55

C e m e n t i n g Te c h n o l o g y

The Effect of Cement Heat of Hydration


on Maximum Annular Temperature
Recent advances in electronics technology
have made it possible to monitor and
record real-time annular temperatures in
wells during and after primary cementing.
These advances have allowed operators to
measure the annular temperature during
the critical period when cement hydration
occurs. Current industry practice is to test
critical properties of set cement at bottomhole static temperature (BHST). If the
short-term maximum annular temperature is significantly different from the
BHST, laboratory tests run on cement at a
steady BHST may be inaccurate.
Introduction
Concurrent with placing annular sealing
cement in ever more challenging conditions,
the oil and gas industry also has begun to
examine more closely the short- and longterm mechanical properties of annular
sealants and the effect these properties have
on long-term annular isolation. To test the
mechanical properties of cement under well
conditions, the cement must be cured or
hydrated for the appropriate time under
temperature and pressure conditions as close
as possible to downhole conditions in the
well. If the cement system is not hydrated at
the correct temperature, mechanical property tests may yield results that are inconsistent with the properties developed by the
cement in the well. Additionally, induced
stresses in a set annular cement sheath have
been shown to be affected by short-term
changes in wellbore temperatures.
Advances in completion technology have
produced such tools as external casing perforating guns that must be sealed in the
cemented annulus. The tools cemented in
the annulus are subjected to the same conditions as the cement that hydrates around
them. Because most tools have certain
design limits for time at temperature and
pressure, knowledge of actual downhole
conditions can be critical.
Most set cement performance testing has
been performed at or very near BHST at the
depth the cement is placed. Typically,
cement curing chambers that hydrate the
cement and hold it for a given time under
downhole conditions before testing will be

AUGUST 2003

heated slowly from a point near the anticipated bottomhole circulating temperature
(BHCT) to BHST. The temperature will be
held at BHST until the end of the curing
process, at which time the mechanical property testing will be performed.
Now, it is recognized that the exothermic
nature of Portland cement hydration can produce significant amounts of thermal energy
once cement hydration begins. This heat flux
can cause the annular temperature to be
greater than the BHST for some period of time.
Data Acquisition
Electrical resonating diaphragms (ERDs)
were used to detect downhole annular temperatures in real time. Reliability is of utmost
importance because the instrumentation is
not retrievable. The equipment can operate
continuously at temperatures as high as
480F. The technology has been field proved
for 8 years in more than 500 systems.
One reason for the high reliability level is
the removal of downhole electronics. The only
downhole sensing components of the ERD are
passive resonators that are excited from the
surface. The sensor is not exposed to pressure,
and its resonant frequency is a function only of
temperature. The passive resonators oscillate
at a frequency dependent on temperature. The
frequency is detected locally and transmitted
electrically to surface by permanent annular
wireline. ERD technology uses a pulse of electricity down an electrical cable to measure the
response of a passive sensor.
Wells
Four wells with different annular configurations were selected with different cementing
systems and slurry density. A cement-density factor (CDF) was defined as the cement
slurry density multiplied by the annular volume of a 1-ft length.
Well 1 was a shallow 2,543-ft vertical well
in Canada. Because the well had multiple
pay zones, the annular temperature sensor
was set at 1,500 ft. In this well, 31/2-in. casing was set in a 77/8-in. hole. The well had a
very low BHST of only 66F. The cement
slurry used was a 15-lbm/gal composite
blend system that had a 30.46-lbm/ft CDF. In
this well, the fluids cool as they are injected.

This article, written by Assistant


Technology Editor Karen Bybee, contains highlights of paper SPE 77756,
The Effect of Cement Heat of Hydration
on the Maximum Annular Temperature
of Oil and Gas Wells, by R.L.
Dillenbeck, SPE, T. Heinold, SPE, and
M.J. Rogers, SPE, BJ Services Co., and
I.G. Mombourquette, SPE, Promore
Engineering Inc., prepared for the 2002
SPE Annual Technical Conference and
Exhibition, San Antonio, Texas, 29
September2 October.
For a limited time, the full-length paper
is available free to SPE members at
www.spe.org/jpt.

Wells 2 and 3 were in Alaska. These two


land wells had a total depth of approximately 7,700 ft measured depth (MD). True vertical depth (TVD) of the wells was 7,400 ft
with annular sensors set at 5,500 ft because
of multiple pay zones. Both wells used
15.7-lbm/gal Class G cement slurries. A
31/2-in. casing was run in an 81/2-in. hole giving a 38.43-lbm/ft CDF. BHST at the sensors
was 113F in both wells. Use of these two
wells provided data from two nearly identical
wells to show if results were repeatable.
Well 4 was in Venezuela. Total depth of this
vertical well was 12,900 ft. The annular temperature sensor was at 12,600 ft, and the BHST
at the sensor was 244F. A 51/2-in. casing was
run in an 81/2-in. hole. A moderately extended 14.0-lbm/gal composite blend cement slurry was used. CDF was 23.99 lbm/ft.
Heat of Hydration (HOH) Testing
An isothermal cement calorimeter was used
in the cement hydration heat testing. The
calorimeter used was designed so that all
heat generated must pass through the walls
of the calorimeter. A constant temperature is
maintained by passing temperature-controlled water through the outer calorimeter
jacket. When heat begins to flow through
the calorimeter walls, a temperature difference is established that is directly proportional to the heat flow. The original calorime-

55

77756.qxd

7/11/03

7:47 AM

Page 56

ter was modified and


equipped with commercially
available heat-flux sensors. A
thin-foil, 40-junction thermopile was bonded to both
sides of the Kapton barrier
with known thermal characteristics. Because the heat flow
is directly proportional to the
temperature differential across
this thermal barrier, an exact
heat flow rate can be calculated and displayed by measuring this difference.
The cement slurry used on
Fig. 1Real-time temperature in Well 2.
each of the four wells was
blended and mixed in the laboratory for HOH testing. Each
slurry was conditioned in an
atmospheric consistometer to
bring it up to curing temperature and then placed in the
HOH test chamber. The test
chamber held approximately
85 g of slurry, and each test was
run for approximately 70 hr.
HOH was determined twice
for each slurry. First, each system was tested with the
calorimeter ambient temperature held constant at 80F to
compare the heat evolution
curves and HOH of the variFig. 2Real-time temperature in Well 3.
ous systems at the same temperature. Next, each cement
system was tested for HOH, but the test cell annular sensor did not allow cement HOH to
in the calorimeter was heated to approxi- be detected, there were no recorded indicamately the same temperature that the curing tions to suggest that this might have
cement would be exposed to in the well. occurred. However, data from testing perBecause of temperature limitations of the formed suggests other possible answers. The
submersible pump used to circulate heated HOH energy yield of the cement slurry
fluid through the calorimeter cell, the maxi- pumped in this well was nearly the lowest of
the four wells examined when compared
mum temperature was limited to 150F.
with the other slurries at 80F. This was
caused by the presence of significant
Discussion
Even though annular temperature data was amounts of extenders and other nonexotherrecorded in both Alaska wells until after the mic materials in the slurry. These materials
peak cement hydration temperature was were used to reduce the system density and
observed, data acquisition was temporarily to increase resistance to gas migration during
terminated to remove the drilling rig from hydration. Although the total HOH for the
the wells. The decline portion of the annular Venezuelan slurry was very similar to the
temperature curve after its peak is a projec- total HOH of the slurry used on the Canadian
tion. Several days after the drilling rigs were well, the heat evolution curve for the
removed, data acquisition resumed, and Venezuelan slurry was flatter with a signifiannular temperature data was used to deter- cantly lower peak indicating that the energy
mine the final stabilized static annular tem- yield was spread over a longer time interval.
perature. On Wells 1 and 4, annular temper- The CDF for this well was the lowest. Even
though this well was the hottest of those
ature data acquisition was not interrupted.
In the deep well in Venezuela, no definitive examined, its 186F average BHCT during
temperature spike greater than the BHST was cementing left a significant temperature
observed. Even though the possibility exists increase of 58F before equilibrium BHST
that an incomplete cement sheath near the was again achieved. It is possible that the lim-

56

ited amount of energy yielded


by the cement in this well
served only to return the annular temperature to BHST more
rapidly than would occur without cement hydration.
On the basis of literature
reviews and previous research,
the authors expected that the
maximum annular temperature during cement hydration
would be noticeable. However,
they did not expect to see the
temperature differential of
more than 50F between the
maximum annular temperature and stabilized static temperature at the sensors on
Wells 2 and 3 shown in Figs. 1
and 2, respectively.
The total amount of heat liberated by each pound of
cement is dependent on the
temperature at which the
cement hydration occurs. The
cement system used in Well 1
generated approximately 42%
more total energy in the 70hour 80F test than it did
under the test in the well at
66F. In Wells 2 and 3 in
Alaska, the HOH tests yielded
70-hour cumulative totals at
BHST that were less than the
70-hour 80F HOH tests.
Although the maximum energy flux from
these slurries was approximately 40% higher under downhole temperature than at
80F, the small slurry sample size allowed
the exhaustion of chemical reactants that
drive the exothermic hydration process long
before the 70-hour test ended.
Conclusions
1. Cement-hydration-caused temperature
increases during cement hydration ranged
from none to 52F.
2. Cement HOH values obtained at ambient temperature are significantly lower than
the heat yield for a slurry when hydration
occurs at downhole temperatures.
3. The annular temperature change that
occurs after the cement achieves initial set
and HOH peaks can induce significant
stresses in the cement sheath.
4. Downhole HOH and CDF play significant roles in determining maximum annular temperature during cement hydration.
5. For low HOH cements with large differences between BHCT and BHST, maximum annular cement temperature may not
JPT
exceed BHST.

AUGUST 2003

Das könnte Ihnen auch gefallen