Vous êtes sur la page 1sur 13

IADC/SPE 87211

Issues and Solutions for Cementing Expandable Liners: A Case History


Rashid Abdul Rahman, SPE, BJ Services Company, Ahmad Humaizy Zulkafly, Petronas Carigali Sdn Bhd

Copyright 2004, IADC/SPE Drilling Conference


subassembly installed at the bottom of the liner. The launcher
This paper was prepared for presentation at the IADC/SPE Drilling Conference held in Dallas, subassembly contains the expansion mandrel that will be used
Texas, U.S.A., 2–4 March 2004.
to permanently deform the liner. An inner workstring is run
This paper was selected for presentation by an IADC/SPE Program Committee following
review of information contained in a proposal submitted by the author(s). Contents of the
and latched into the top of the mandrel. Cement slurry can
paper, as presented, have not been reviewed by the International Association of Drilling then be pumped through the inner workstring into the annulus.
Contractors or Society of Petroleum Engineers and are subject to correction by the author(s).
The material, as presented, does not necessarily reflect any position of the International Later, a latching dart is pumped behind the cement slurry to
Association of Drilling Contractors or Society of Petroleum Engineers, their officers, or
members. Papers presented at IADC/SPE meetings are subject to publication review by
seal off the float equipment below the launcher subassembly
Editorial Committees of the International Association of Drilling Contractors and Society of to create a pressure chamber between the shoe and the
Petroleum Engineers. Electronic reproduction, distribution, or storage of any part of this paper
for commercial purposes without the written consent of the International Association of Drilling expansion mandrel in the launcher subassembly. Pressure is
Contractors and Society of Petroleum Engineers is prohibited. Permission to reproduce in print then increased in the pressure chamber to first break burst
is restricted to a proposal of not more than 300 words; illustrations may not be copied. The
proposal must contain conspicuous acknowledgment of where and by whom the paper was disks and then transmited to the base of the expansion
presented. Write Librarian, SPE, P.O. Box 833836, Richardson, TX 75083-3836, U.S.A.,
fax 01-972-952-9435. mandrel. The pressure drives the mandrel up to expand the
liner. The inner workstring is continually pulled out of hole as
the mandrel moves up, with each stand of the workstring is
Abstract
broken out and racked back before next section of liner
The oil and gas industry has in general already in place an
is expanded1, 2, 3.
established set of good cementing practices to meet the overall
objectives of cementing. These practices have been refined
Good Cementing Practices
and perfected over decades of industry experience in the
Decades of cementing experience taught the industry to
discipline. However, new technology of expandable liners
establish good cementing practices to be followed to achieve
brought about its own set of cementing issues and
the overall objectives of cementing. This has been well
requirements based on the equipment design constraints. Some
documented in numerous literatures available to everyone
of the requirements run counter to the established good
interested. The authors will attempt to cover briefly only on
cementing practices. This paper discusses some of the new
the relevant practices here for the benefit of relating them to
cementing issues and requirements for expandable liners and
the requirements for cementing expandable liners. Thus, it is
their implications on the overall effectiveness of the cement
not the intention of the authors to deal with all the established
job. Optimizing the cement job involves addressing the issues
practices here. The practices that will be considered are only
and taking a compromise between the new cementing
the following main issues:
requirements and the established good cementing practices.
This paper discusses some of the available solutions for the
cement job optimization with a view to meet the overall • Well conditioning
cementing objectives. Finally, a case history where the above • Pipe centralization
has been successfully implemented is presented and discussed. • Mud removal
• Cement slurry design
Introduction
Expandable liner is one of those new technologies that has Well Conditioning
allowed the industry to expand the drilling operating The first thing to consider is the wellbore geometry. Assuming
envelopes. Its main advantages are to allow the preservation of that the driller has done all that is possible to ensure the hole
larger wellbore sizes and case off troublesome zones such as contains as fewer washouts as possible and that it is as near
overpressured and lost circulation zones. Installing expandable gauge as possible, it is important to determine the hole size
liner starts with drilling an overgauged hole. The overgauged accurately. This can be done by using a 4-arm open hole
hole allows the liner to be run through a casing of certain size, mechanical caliper and pumping fluorescent dye to further
expanded and cemented, while still preserving the wellbore confirm the measurement with the caliper.
size large enough to run the casing regularly run through the
previous casing. For example, installing an expandable liner Next, one must consider the mud conditioning. Significant
through a 9-5/8” casing would still allow the operator to run change to the mud rheology should be performed during the
the 7” liner or casing through the expandable liner. After the last trip of drillpipe prior to running the casing. Mud needs to
open hole is drilled, the expandable liner is run with a launcher be circulated before removing the drillpipe. Otherwise,
unconditioned mud may have sufficient time to gel during the
2 IADC/SPE 87211

period when drillpipe is removed, logging and running To further improve the pipe centralization, it is recommended
casing4. After the casing is landed to TD, the mud should that a self-centralized float shoe and collar be used to ensure
again be circulated to remove entrained gas and cuttings, break good cement coverage around the shoe joint. The effectiveness
gelled mud and lower its gel strength, yield stress and plastic of the equipment is greatly enhanced with its up-jet flow
viscosity. Measurements of the mud’s 10-minute gel strength profile at the side of the float shoe, in addition to the standard
should continuously be taken to confirm that it is kept low to downward flow. It is reported in various publications that the
allow efficient mud removal. Along with volume of mud equipment significantly reduced the chances of having to
filtrate, the 10-minute gel strength greatly influences the mud perform squeeze jobs.
mobility factor (MMF), which critically determines the
percentage of mud removed from the annulus5. Both the Mud Removal
volume of mud filtrate and 10-minute gel strength need to be Assuming that the necessary well conditioning and pipe
reduced to maximize MMF. centralization have been performed, the mud removal process
prior to cementing can be optimized by ensuring that the
Next, the hole should be cleaned from cuttings especially in following practices are followed. It is a simple concept to
the build and horizontal sections, where applicable. To achieve grasp that to get a good cement coverage and bond around the
this, studies indicate that the mud annular velocity should be pipe in the annulus, the existing mud needs to be displaced for
kept above 262.5 ft/min6 or a circulation rate of at least 500 the cement to take its place. To achieve this goal, the first
gal/min should be maintained7. Also, it was concluded that thing to do is to optimize the chemical wash and spacer.
there appears to be a threshold value of yield points below
which a continuous solids channels will appear. The minimum Chemical wash and spacer should always precede the cement
yield point values of 20 lbf/100ft2 and 28 lbf/100ft2 are slurry11. Ideally, all fluids should be in turbulent flow,
required for 60° and 85° deviation respectively7. Some studies including the mud. However, it is normally difficult to achieve
indicated that at high angle, a minimum of 50 lbf/100ft2 is turbulent flow with the cement slurry and the weighted spacer.
actually required to properly remove drill cuttings8. Thus, it is So, turbulent flow can instead be achieved with the chemical
essential that the mud properties be monitored and controlled preflush. The weighted spacer and cement slurry can then
to meet these criteria prior to cementing operations. easily displace the thinner fluids (mud and chemical wash),
which have low resistance to flow. In addition to turbulent
Lastly, mud circulation should be at least “bottoms up” but flow, other factors to consider in chemical wash and spacer
preferably two hole volumes. It should be continued until a design are the required volume and its chemical properties.
minimum of 95% of the circulatable mud has been moved9. Numerous studies have indicated that the more spacer volumes
Turbulent flow should be maintained provided formation used, the better chances of optimized mud displacement being
breakdown pressures are not exceeded. achieved12 while others have shown the opposite result13,14.
However, the actual recommended volumes to be used are
Pipe Centralization inconclusive with some recommending 4 minutes contact
Centralization is essential to provide an annulus with open time, while others recommending 10 minutes as the
flow paths. If the casing is not centralized, the pipe will trap minimum4. Still others recommended 8 minutes contact time15.
mud against the low side of the hole. Because of the In any case, the chemical wash and weighted spacer must be
differences in the flow path, there is no flow regime or tested for compatibility with the cement slurry and mud.
practical flow rate that can remove the trapped mud. Studies
concluded that when the API casing standoff was less than Based on the authors’ experience, volumes of 20 bbls of
60%, mud contamination and significantly poor displacement chemical wash and 80 bbls of weighted spacer have
efficiency prevailed despite careful control of mud, preflush consistently worked well in 9-5/8” casing and 7” liner jobs in
and cement slurry properties10. Although API recommends a deviated wells. The 20 bbls chemical wash is essentially water
67% casing standoff, it is recommended that a standoff of at mixed with strong water wetting surfactant. It is intended to
least 80% be designed, because the actual standoff will break and/or thin the mud and mud filter cake in the annulus
normally turn out to be less than the designed value as the with its turbulent flow regime. Because of its Newtonian fluid
centralizers embed to some extent into the formation. nature, turbulent flow is easily achieved with chemical wash.
An alternative to chemical wash when oil-based mud is used is
Centralization is difficult in highly deviated wells. To to precede the weighted spacer with some volume of base oil
maintain optimum standoff, it is recommended to keep the to help thin the mud and soften the mud filter cake. From the
spacing between the centralizers below 20 ft or 6.1 m9. Solid authors’ experience, pumping 10 – 20 bbls of base oil before
body spiral centralizers are recommended because they give pumping the weighted spacer yields very positive results.
positive standoff even when extreme loads are placed on the
centralizers. The centralizers should be allowed to move freely The weighted spacer pumped behind the chemical wash or
between 2 stop rings and not be fixed onto the casing with set base oil is intended to displace the already thinned mud out of
screws. Fixing the centralizers onto the casing will cause the the annulus by the combined effect of density and yield
centralizers to deteriorate because of wear during the rotation strength differentials. The spacer is weighted to 0.5 ppg
and reciprocation of the pipe. heavier than the mud to impart buoyancy effect that will lift
the mud out of the annulus. To weigh up the spacer, it has to
be gelled up with polymer to provide the solids suspending
IADC/SPE 87211 3

properties. The yield strength differential between the spacer mud16. It was shown that where this condition is satisfied, the
and thinned mud is designed to provide a flat interface flattened form of the velocity profile of the slurry greatly
between the mud and spacer to enhance displacement enhances the mud removal even when the standoff is only
efficiency further16,17,18. It is also important to prevent 46%17. However, the slurry yield point value should ideally
channeling of the spacer through the mud or intermixing of the not exceed 30 lbf/100ft2 because excessively high cement
two fluids, resulting in poor displacement efficiency. In slurry yield point can increase the average velocity required to
addition to providing yield strength differential, the weighted initiate flow in the narrowest part of the eccentric annulus.
spacer also contains weighting agent (barite) that enhances the This will cause the cement slurry to be diverted away from
turbulent stresses and erosion on the mud. This further that part and left it uncemented.
improves the mud displacement efficiency to prepare for the
cement slurry placement. Lastly, the cement slurry should be Third, cement slurry must have low fluid loss, among other
pumped as fast as possible even when turbulent flow cannot be reasons, to prevent premature dehydration while flowing
achieved because displacing at high flow rates is more across permeable zones. This is especially important when
effective than plug flow displacement19. cementing highly deviated sections or horizontal sections
where the exposure to permeable zones is typically bigger. For
It is important that the pipe be either rotated or reciprocated this purpose, an API fluid loss level of 50 cc/30min/1000psi is
during the pumping of the preflush and spacer to aid in the recommended. Excessive fluid or cement filtrate loss into the
breaking of the mud gel strength and filter cake. A rotation formation will alter the properties of the cement slurry from its
rate of 10 - 30 RPM or reciprocation rate of 20 – 40 ft in 1 to 5 designed properties21. Thus, actual thickening time, rheology,
minutes is generally recommended. mechanical strengths and gas tightness might not be sufficient
to ensure quality cement is placed in the annulus. In addition,
Cement Slurry Design to achieve a gas tight cement slurry while the slurry is losing
To achieve full zonal isolation especially in highly deviated its hydrostatic pressure during cement hydration and setting,
sections, the cement slurry must be designed to have several the API fluid loss level needs to be limited to a maximum of
essential characteristics. First, it must have zero free water in 50 cc/30min/1000psi. This is to ensure that the slurry does not
both API 90° and 45° free water tests. This is to prevent the prematurely lose its hydrostatic pressure before the hydration
creation of channels at the upper side of the annulus that will process takes place. The hydrostatic pressure is important to
allow fluid crossflow. prevent formation fluids or hydrocarbon from entering the
wellbore and migrating to the surface.
Second, the cement slurry must have good rheology to be able
to flow through narrow annulus at the bottom side of the Cementing Requirements and Issues for
annulus. For a Bingham Plastic fluid, this means that the Expandable Liners
plastic viscosity and yield point values of the slurry must not Novel technology like expandable liners imposes new set
be excessively high. A plastic viscosity of less than 100 cp and of cementing requirements based on its design constraints.
yield point of less than 30 lbf/100ft2 are normally desired, Among the new requirements are22:
depending on the yield point value of the mud. It was
previously thought that cement slurry should have very low • The cement slurry requires additional extended fluidity
viscosity to allow the achievement of turbulent flow as it is time for liner expansion operation in addition to the
being pumped into the annulus. This was thought to be helpful thickening time for cement slurry mixing and placement
in the process of mud displacement. However, many studies • The slurry pump rate should not exceed 3 or 4 bpm
concluded that it is virtually impossible to achieve turbulent • No centralizers will be used
flow with cement slurry in most well conditions at the typical • The slurry volume to be pumped is based on post-
rheological properties of most cement slurries. expansion liner diameter
• Pipe movement is not possible
In addition, it was also proven that it is difficult to accurately • Cement slurry must have excellent stability for extended
determine the required critical rates at which turbulent flow period while in static condition
can be achieved because of the phenomenon called “slippage”
of the cement slurries at the pipe surface20. So, instead of Extended Fluidity Time
trying to design the slurry to have the rheology to achieve Typically, a 10 – 12 hour fluidity time is required to allow
turbulent flow, it is recommended to design it to enhance the sufficient time for the expansion operation and the expansion
displacement efficiency of the mud via yield strength assembly to be retrieved. Fluidity time is defined as the time
differential17. It was shown that poor mud displacement can be the static gel strength is below 100 lbf/100ft2 (Ref. 23). This
seen in an eccentric annulus when the mud is displaced with time is generally taken to be about twice the time for the actual
cement slurry which has the same rheological properties as the operations to allow for corrective actions in the event of any
mud15. To enhance the mud displacement efficiency, the operational problems. The nature of the cement slurry does not
plastic viscosity and yield strength of the slurry should be favour a lengthy thickening time while maintaining the
designed to be higher than those of the mud, with the yield designed properties of the cement slurry. This is expecially
strength differential showing more significant effect. It was true when the slurry is left in static condition at high
suggested that where possible that the yield point of the temperatures. Most fluid loss additives, extenders and free
cement slurry is at least 10 lbf/100ft2 higher than that of the
4 IADC/SPE 87211

water control additives rely on one type of suspending agent or retrieved. This issue can be addressed with accurate
the other to keep the cement particles and other solids determination of the hole size with the proper tools.
suspending in the slurry. They are also needed to tie up the
water by various mechanisms to prevent the formation of free Pipe Movement is Not Possible
water and channels at the top of the cement sheath. In most Pipe rotation and reciprocation are not generally possible
cases, these additives work well within the short period that is because it may cause damage to the expandable liner at the
typical of the most required thickening times. These range previous casing shoe. The elastomers fitted onto the
from 1 to 6 hours. However, for any period exceeding 10 expandable liner at the overlap section may get damaged
hours, the performance of these additives typically deteriorates during rotation and reciprocation operations if they contact the
significantly. The slurry will start to experience severe previous casing shoe. This problem can also be minimized by
gelation, sedimentation and density segregation. Free water the proper designs of the fluid rheological properties.
will also start to appear and thickening time starts to become
unpredictable. The end result would be a poor quality cement Excellent Slurry Stability in Extended Period
sheath across the expandable liner. This problem can however Cement slurry needs to have excellent stability during the
be overcome with novel cement slurry design technique to be extended time while the expandable liner is being expanded.
discussed later. There must not be any solids settling which results in density
segregation during this period. The slurry also needs to
Low Pump Rate maintain its fluidity to ensure the expansion operation can be
During the displacement of cement slurry in cementing carried out smoothly. If the slurry loses its fluidity that is
expandable liner operation, the pump rate needs be kept low to manifested with an increase of static gel strength above 100
prevent the cement slurry from lifting or displacing the lbf/100ft2, proper expansion may not be achieved. The highly
cuttings and gelled mud filter cake out of the annulus to the gelled slurry may induce losses into the formation during the
overlap section at the top of the liner. In the undesirable event expansion operation. In addition, instead of the top of cement
that the cuttings and cakes were lifted to the overlap section, (TOC) level rises to the designed depth, it may remain at the
they could potentially fall into the liner, causing problems to pre-expansion level because the highly gelled slurry is
retrieve the expansion assembly during the expansion predominantly being squeezed into the formation instead
operation. Also, the cuttings could potentially settle around the displaced upward in the annulus. This could result in failure to
overlap section, preventing the full expansion of the isolate parts of the problematic zone. This problem can be
expandable liner and preventing the achievement of hydraulic managed with a new cement slurry design technique to be
seal at the overlap. In addition, low pumping rate is also discussed later.
required to prevent from damaging the elastomers at the
overlap from shearing action with the abrasive cement. Good Solutions for Cementing Expandable Liners
mud removal is especially difficult to achieve in cementing To ensure the cementing objectives are met while working
expandable liner because of the tight annular space. A with the constraints imposed by the expandable liner
combination of low displacement rate and tight annular space equipment, a two-pronged strategy can be employed. First, the
would lead to a poor quality of cement sheath behind the remaining good cementing practices that can still be employed
expandable liner. This problem will need to be minimized by should be implemented to full effect. Second, a special cement
the proper designs of the fluid rheological properties. slurry system providing enhanced relevant properties can be
used to offset the negative effects of the expandable liner
No Centralizer equipment constraints.
It is not possible to utilize centralizers on expandable liner
because of the tight annular clearance between previous casing Based on the above strategy, the following practices will be
shoe and the expandable liner. Thus, depending on the well implemented to their full effect without any compromises:
deviation, the pipe standoff will range from bad to non-
existent. This would result in poor cement coverage across and • Open hole size must be determined as accurately as
around the expandable liner. Similar to the above, this issue possible with 4-arm caliper and fluorescent dye.
can be partly alleviated by the proper designs of the fluid • Mud must be circulated before removing the drillpipe.
rheological properties. • At least 2 hole volumes of mud must be circulated. Use
carbide-pill method to determine if enough mud has
Slurry Volume Based on Liner Post-expansion OD been circulated.
The expandable liner is expanded after the cement slurry is • Mud circulation must be at a rate of at least 500 gal/min.
placed in the annulus. Since the post-expansion OD of the • During mud conditioning, mud 10-minute gel strength
expandable liner is around 11 – 16% greater than the pre- must be monitored to ensure it keeps decreasing.
expansion OD, the top of cement level will rise when the liner • During mud conditioning, the yield point value must not
is expanded. Thus, pumping only the annular volume based on be lower than 20 lbf/100ft2 to ensure proper
the post-expansion OD will ensure that the TOC will not rise solids suspension.
to the overlap section and overflow into the liner. If the
• A chemical wash must be pumped ahead of
cement slurry overflows into the liner, it could prove
weighted spacer.
disastrous if it sets before the expansion assembly could be
IADC/SPE 87211 5

• Weighted spacer must have density of 0.5 ppg higher than resisting degradation over time while in such
the mud to impart buoyancy effect. harsh environments.
• Total contact time for preflushes must be at least
8 minutes. A Case History
• Weighted spacer yield point must be designed to be in The above solutions have successfully been used in various
between those of the mud and the cement slurry to ensure expandable liner cementing jobs performed by one of the
flat velocity profiles are obtained during displacement. author’s employer worldwide. In Malaysia, one job was
• The cement slurry rheological properties must not be the performed in November 2002 in a high temperature high
same as those of the mud. Its Pv and Ty must be higher pressure (HTHP) appraisal well Bergading Deep-3, while
than those of the mud. another job was just performed recently in another similar
• The cement slurry yield point value must be at least 10 appraisal well. The following information was captured from
lbf/100ft2 above that of the mud to ensure flat velocity the job performed last year.
profile is obtained at the mud-cement interface.
Well Information
Effective combined actions of chemical wash and weighted The well Bergading Deep-3 appraisal well is located in the
spacer can maximize the removal of cuttings out of the Block PM-302 offshore Peninsular Malaysia approximately
annulus prior to the cement placement operation. This helps 320 km northeast of Kemaman Supply Base (Refer to Fig. 1).
reduce the chances of having the cement slurry lifting and The objectives of drilling the well were to firm up
depositing the cuttings at the overlap section. hydrocarbon reserves in the northern Malay Basin and to test
well deliverability through production tests to the primary
Next, with regards to the cement slurry, it must be designed to target one each in lower sand packages. It was classified as
meet the following criteria: HPHT well based on the prognosed bottomhole static pressure
(BHSP) and BHST correlated from the offset wells. The
• The cement sheath must have at least 500 psi 24-hr undisturbed static bottom hole temperature at the total depth of
compressive strength to minimize wait-on-cement (WOC) 2975 m-TVDDF was expected to be 330 °F while the
time for drilling ahead. maximum anticipated pore pressure was around 8860 psi. The
• The cement slurry must have higher ductility to cope with critical part is to look for the onset of overpressure zone that
downhole stresses in the narrow annulus. This is done by was expected to be at 1794 m TVDDF. The mild increase in
slightly lowering the density of the cement. pore gradient from 0.572 psi/ft to 0.606 psi/ft within 391 m
• Cement slurry must be designed to have zero free water. interval had triggered the idea of setting the 9-5/8” casing to
cover up this overpressure zone down to 2225 m TVDDF in
• Cement slurry must have API fluid loss of less than 50 cc.
which the increasing trend softened.
• Cement slurry must have a gel strength of less than 100
lbf/100ft2 during the thickening and fluidity times.
The actual bottom hole data obtained from wireline logging
• Cement slurry must have excellent rheology (low plastic results showed slightly different figures from the prognosed
viscosity and moderate yield point) to allow it to be
data in that the BHST turned out to be 323°F at TD and the
placed uniformly in the eccentric annulus.
BHSP was approximately 6822 psi. Further data had shown
• Cement slurry must have excellent long-term stability (no that the onset of overpressure ramp was first observed at 1820
sedimentation and density segregation). m-TVDDF. Since the 9-5/8” casing was set 133 m off the
• Cement slurry must have higher resistance to mud original setting depth at 2225 m-TVDDF, the 7-5/8”
contamination because of the increased likelihood of expandable casing came into the picture to save the larger
intermixing with mud at low displacement rate. casing string. In this case, solid expandable casing basically
acts as a steel mud cake to cater for the loss circulation zone.
This second strategy can be achieved with the use of a special
cement slurry that is designed to be storable for extended The dominant lithological profile for this interval was
period in fluid state. The special slurry utilizes a patented consisted of sandstone section with interbedded claystone and
multipurpose additive incorporating iota carrageenan stringer of siltstone (Refer to Fig. 2 for more lithological
suspending agent to provide the necessary stability24. This information). A thin coal section located at 2450 m-TVDDF
additive, when used in combination with a gluconic acid based was believed to be the weak point where the loss circulation
retarder, produced exceptionally stable slurry that was superior zone occurred during drilling this interval.
to any previously known suspending agents. Unlike regular
cement slurries that employed previously known suspending Drilling Information
agents, the special cement slurry does not exhibit solids Fig. 3 shows the casing design scheme for the well. Drilling
settling, density segregation, formation of free water and high the 8-1/2” x 9-7/8” hole section to accommodate the setting of
gelation that are caused by the deterioration of the suspending expandable casing was a real hassle in the hard formation. The
agent during the extended fluidity time. The slurry also hole was basically in good condition but due to the failures of
exhibited better resistance to harsh chemical environments near bit reamers to perform, logging results showed the hole
such as high pH and high calcium ion content, which are was not opened to the desired hole size to accommodate the 7-
typical in Portland cements. All these properties are borne out 5/8” expandable liner. The hole had to be redressed by
by the robustness of the iota carregeenan suspending agent in
6 IADC/SPE 87211

additional 3 bits and near bit reamers run after the hole had rise to the overlap section and overflow into the expandable
been found under gauged for 7-5/8” expandable liner liner. The job was actually performed twice by no fault of the
preparation. Drilling this section from 2108 m-TVDDF to cement. The first time it was performed, there was a leak in
2527 m-TVDDF indicated the mild-increasing trend of the expandable liner system, preventing it from being
overpressure section over depth. The calculated pore pressure pressured up prior to expansion operation. The cement slurry
gradient was 0.78 ppg/100 m with the projected pore pressure had to be recirculated out of hole and the whole expandable
at 2527 m-TVDDF was 13.6 ppg. liner assembly retrieved. This was achieved without problems,
because of the stability of the cement slurry. It was discovered
The 8-1/2” x 9-7/8” hole section TD was called at 2527 m that a bolt had become embedded into the expandable liner
TVDDF due to losses experienced at a rate of 60 bbls/hr at aluminium shoe, preventing tight sealing between the shoe and
645 GPM and 30 bbls/hr at 500 GPM using a mud weight of wiper plug.
14.4 ppg. At a flowrate of 645 GPM, ECD is 15.3 ppg and at
400 GPM ECD is 15.23 ppg. The mud weight in hole during The second job was performed smoothly. 10 bbls of chemical
losses was maximum 14.4 ppg with the (leak off test) LOT at wash and 20 bbls of 14.8 ppg weighted spacer were pumped at
9-5/8” shoe was 15.4 ppg equivalent mud weight (EMW). 3 BPM to displace 14.3 ppg OBM. The weighted spacer
Mud losses stopped after 100 bbls of 14.6 ppg LCM pill was rheological properties were Pv at 32 cp and Ty at 23
spotted on bottom at a flowrate of 100 GPM. The bottomhole lbf/100ft2, which were designed to be in between those of the
circulating temperature (BHCT) was 266°F and bottomhole mud and the slurry. During this time, the surface pressure
static temperature (BHST) was 300°F. hovered between 600 – 700 psi. 33 bbls of batch-mixed 15.0
ppg gas tight special cement slurry was later pumped at 3
Job Information BPM with the surface pressure kept between 600 – 700 psi.
Based on the experience in a prior exploration well drilled 3.5 bbls of chemical wash was then pumped behind the
close by, it was decided to utilize the expandable liner to case cement slurry before the latching dart was dropped. Later,
off the lost circulation zone in the 12-1/4” section This is to 125.8 bbls 14.3 ppg OBM was pumped at 4 BPM to displace
allow the operator to drill ahead without losing hole size. The the slurry with the cement unit. During the displacement, the
operator had experienced total losses in the prior well that surface pressure remained between 600 – 900 psi, until when
resulted in massive volume of oil-based mud lost. the latching dart bumped into the float equipment. At this
time, the pressure instantly increased from 700 psi to 1200 psi,
For the current well, the expandable liner was used to case off stabilized at 1200 psi for 1 minute, before increasing again to
the lost circulation zone so that drilling ahead with heavier 3000 psi within 2 minutes. Displacement was continued for
mud was possible. A 7-5/8” expandable liner was set in the 8- another 18 minutes with the pressure slowly increasing to
1/2” open hole section underreamed to 10” at TD 2527 m- 3300 psi, when the burst disks broken the pressure dropped
TVDDF. The liner was hung inside 9-5/8” 47 ppf casing set at back to zero. Full returns were observed throughout the job,
TD 2092 m-TVDDF by way of an overlap between the but there were no cement returns. The pipe was not
expandable liner and the 9-5/8” casing. The expandable liner reciprocated or rotated during the job. The job took 90 minutes
has sets of elastomers on its OD at the 60 m overlap section to from the start of the weighted spacer was pumped to the
give full hydraulic seal and act as liner top packer. The breaking of the burst disks. The rest of the operations involved
expandable liner has an OD of 7-5/8” and ID of 6.875” in pre- the expansion of the liner that took 220 minutes to complete.
expanded form. In post-expanded form, the OD is 8.544” and
ID is 7.830”. To expand the liner, it needed to be pressured up to about
3500 psi. Pumping at an average 0.75 – 1.00 BPM, the
Prior to pumping the cement slurry, two hole volumes of mud workstring pressure was slowly increased to between 3200 –
were circulated at 480 gal/min. This pumping rate was slightly 3500 psi. The liner was expanded by slowly pulling out of
less than the recommended practice of 500 gal/min, but it was hole the expansion assembly with pick up weights of between
acceptably close enough to the target minimum rate and since 180 – 200 kips. The pressure had to be bled off after
it could not be increased further to avoid inducing losses. The expanding each stand to rack the pipe. Expanding the 60 m
mud rheological properties at the completion of the mud overlap section which acted as the liner hanger required 290 –
circulation were Pv at 33 cp, Ty at 19 lbf/100ft2 and its 10- 300 kips. 485 m of expandable liner from 2517 – 2032 m was
minute gel strength at between 11 - 13 lbf/100ft2. These values expanded. Once the expansion was completed, contaminated
are in accordance to the designed values to optimize mud mud was reverse-circulated.
removal. As desired, the mud yield point value was less than
the cement slurry yield point value by 10 lbf/100ft2. In Later, once the well was drilled to TD, a 7” liner was run and
addition, the mud 10-minute gel strength was maintained at set through the expandable liner and hung off the 9-5/8”
low values. casing above the overlap section of expandable liner/9-5/8”
casing. The 7” liner was cemented with regular 15.8 ppg gas
A 4-arm caliper log was run to determine the open hole size tight cement slurry. CBL was run to determine the cement
OD at an average of 9.97”. Using this OD, the cement slurry quality behind the 7” liner. It showed unsatisfactory cement
volume required was decided to be at 75% of the pre- quality in some sections of the 7” liner. CBL was run again a
expansion annular volume to ensure that the TOC would not few days later, but this time 500 psi pressure was applied in
the liner. The CBL showed improved cement bond and the
IADC/SPE 87211 7

operator readily decided to production test three zones across produced with lower density slurry requires less tensile
the expandable liner section, putting the cement quality behind strength to avoid mechanical failure than that of cement sheath
the expandable liner to the test. The production tests were produced with higher density. The resulting 24-hour
conducted successfully without any crossflow between the compressive strength of 600 psi was slightly lower than
zones, proving that zonal isolation was achieved behind the normal, but well within the operator’s comfort level of 500 psi
expandable liner and the 7” liner, even though CBL was not it required for drilling ahead.
optimum. It is interesting to note that the expandable liner was To determine the thickening time of the slurry, the bottomhole
the first installed in open hole in Malaysia. circulating temperature (BHCT) was used. This was done as
per normal API procedure. However, to determine the slurry
Cement Slurry Information properties during its fluidity state, the bottomhole static
The cement slurry formulation is as follows and its properties temperature (BHST) was used to simulate the slurry in static
are as per Tables 1, 2 and 3: condition after it was placed in the annulus. The consistomer
meter motor was stopped and it was only turned on after 2
API Class G cement hours, 4 hours, 6 hours and 10 hours to check the Bearden
+ 35% BWOC silica flour units at those times. In a way, this method ensured that the
+ 0.03 gps silicone defoamer slurry was competent even at the worst-case scenario. The
+ 0.75 gps PEI-based fluid loss additive Bearden unit readings at these intervals are as per Table 4.
+ 1.00 gps gluconic acid-based retarder These readings showed that the slurry was remarkably stable
+ 0.30 gps PNS-based dispersant and had low gel strength over the fluidity time.
+ 0.35 gps patented multipurpose additive
+ 5.80 gps seawater 2-hours 4-hours 6-hours 10-hours
4 6 7 6
Properties Values
Density 15.0 ppg Table 4 – Slurry Bearden Unit Readings at 300°F
Thickening Time at BHCT (100Bc) 210 minutes
Fluidity Time at BHST (100 Bc) 615 minutes As mentioned before, to provide the necessary stability to the
API Fluid Loss at BHCT 46 cc/30min/1000psi slurry, the patented iota carrageenan suspending agent-based
API Fluid Loss at BHST 72 cc/30min/1000psi multipurpose was used in combination with a gluconic acid
90° API Free Water at BHST Nil based retarder. The high performance of the slurry is
45° API Free Water at BHST Nil illustrated by the above consistent and stable Bearden Unit
24-hr Comp. Strength at BHST 600 psi readings.
36-hr Comp. Strength at BHST 1400 psi
To evaluate the higher contamination resistance of the slurry,
Pv at BHCT 63 cp
the slurry was mixed with small percentages of mud and then
Ty at BHCT 29 lbf/100ft2
cured at BHST to check the compressive strengths. This was
compared with a conventional 15.0 ppg cement slurry that has
Table 1 – Cement Slurry Properties
comparable properties. The compressive strengths were
measured from the time the slurries were first mixed. The
600 300 200 100 6 3
results obtained were as per Table 5. These results illustrated
140 92 73 50 9 6 the higher resistance to contamination property of the special
slurry as compared to the regular cement.
Table 2 – Fann 35 Readings for Slurry at 300°F
85%:15% 85%:15% 75%:25% 75%:25%
100% 100%
10- 10- 10- Special Regular
Special Regular Special Regular
2-hours 4-hours 6-hours Cement: Cement: Cement: Cement:
seconds minutes hours Cement Cement
Mud Mud Mud Mud
7 27 31 44 52 75 1400 1200 720 445 640 240

Table 3 –Slurry Gel Strengths at 300°F (in lbf/100ft2) Table 5 – 36-hour Comp. Strength at 300°F (in psi)

The slurry density was set at 15.0 ppg because of several Note that the compressive strengths are lower because of the
reasons.. First, the lower density was determined based on the lower density used.
maximum allowable equivalent circulating density (ECD)
generated as per the cementing job computer simulation to To evaluate the gel strength development of the slurry, the
avoid inducing losses during the cement placement. Second, following simple procedure was used28:
the narrow annulus necessitated a robust cement sheath to • Mix slurry as per API Spec 10B29 procedure.
prevent cracking and debonding. By lowering the cement • Place the slurry in HTHP consistometer to simulate
density, the ductility of the cement sheath was improved, placement in annulus at BHCT. Keep it for 3:30 hr:min.
allowing the cement sheath better able to withstand downhole • Remove the slurry from the consistometer and pour it into
stressed acting on it25. References 26 & 27 showed that under the Fann sample cup.
the same applied downhole stresses, the cement sheath
8 IADC/SPE 87211

• After running the standard rheology test, shift to the 600 OBM = oil based mud
RPM setting and stir for 60 seconds. OD = outside diameter
• After 60 seconds at 600 RPM, stop and allow the slurry to ID = inside diameter
remain static for 10 seconds. Switch the Fann to 3 RPM. CBL = cement bond log
• Turn the Fann on at the 3 RPM setting. Observe the bob BWOC = by weight of cement
dial; the reading will rise steadily and then "break back". gps = gallon per sack
The highest reading prior to this break is the gel strength. EMW = equivalent mud weight
• Allow the slurry to rest undisturbed for 10 minutes, and ECD = equivalent circulating density
then repeat step four. BPM = barrel per minute
• Repeat for 2, 4 and 10-hour intervals. RPM = revolution per minute
PEI = polyethylene imine
The gel strength values over these intervals showed that the PNS = polynaphthalene sulfonate
gel strength of the slurry while in static condition remained
stable and low. These values satisfied the requirement of their Acknowledgements
being less than 100 lbf/100ft2 throughout the fluidity time. The authors wish to thank the management of BJ Services
and Petronas Carigali Sdn Bhd for the permission to prepare
To check the tendency for particle sedimentation and density and publish this paper. Appreciation is also extended to all
segregation, the slurry was tested according to the BP Settling individuals who assisted in the laboratory design, engineering
Test procedure. This procedure is included in the Appendix. and field applications of this new method.
The results obtained are as follows:
References
Density: 15.0 ppg (1.80 SG) 1. Enventure Global Technology: “Solid Expandable
Test Temperature: 266°F Tubulars are Enabling Technology”, Drilling Contractor
Test Pressure: 3000 psi (March/April 2001).
Settlement tube length: 203 mm 2. Dupal, K.K., Campo, D.B., Lofton, J.E., Weisinger, D.,
Depth to firm Cement (settlement): 0.0 mm Cook, R.L., Bullock, M.D., Grant, T.P. and York, P.L.:
Density of column, top-bottom: “Solid Expandable Tubular Technology – A Year of Case
1.80,1.80,1.80,1.80,1.80,1.81 SG Histories in the Drilling Environment,” paper SPE/IADC
Nature of free fluid: None 67770, 2001.
3. Filippov, A., Mack, R., Cook, L., York, P., Ring, L. and
The results showed that the slurry was extremely stable during McCoy, T.: “Expandable Tubular Solutions,” paper SPE
the 16 hours duration of the test. The test was performed at 56500, 1999.
BHCT to comply with the standard procedure of the BP 4. Guillot, D., Hendriks, H., Callet, F. and Vidick, B.: “Mud
Settling Test. Removal,” Well Cementing (Developments in Petroleum
Science, 28), Elsevier Science Ltd. (1990), 5-4.
Conclusions 5. Haut, R.C. and Crook, R.J.: “Laboratory Investigation of
1. Optimizing the expandable liner cementing involves Lightweight, Low-Viscosity Cementing Spacer Fluids,”,
taking a compromise between the equipment design paper SPE 10315, 1981.
constraints and the established good cementing practices. 6. Smith, T.R.: “Cementing Displacement Practices – Field
2. Even when the established good cementing practices were Applications,” paper SPE 18617, 1990.
not fully adhered to in cementing expandable liner, the 7. Keller, S.R., Crook, R.J., Haut, R.C. and Kulakofsky,
overall objectives of cementing can still be achieved. D.S.: “Deviated Wellbore Cementing: Part 1 - Problems,”
3. To minimize the negative effects of the inability to JPT, (Aug. 1987) 955-960.
execute some of the critical good cementing practices, the 8. Ravi, K. and Webber, L.: “Drill-Cutting Removal in a
remaining practices must be implemented to full effect. Horizontal Wellbore for Cementing”, IADC/SPE 35081,
3. Special blend of cement slurry provides extremely stable 1996.
properties to satisfy the stringent requirements for 9. Cooper, R.E. and Birch, R.: “Horizontal Well
expandable liner cementing operations Cementing,” in: E.B. Nelson (ed.), Well Cementing
5. Optimizing the expandable liner cementing involves fine- (Developments in Petroleum Science, 28), Elsevier
tuning the cement slurry testing procedures to simulate Science Ltd. (1990), 15-12.
actual operations and downhole conditions. 10. Wilson, M.A. and Sabins, F.L.: “A Laboratory
Investigation of Cementing Horizontal Wells,” SPEDE
Nomenclature (Sept. 1988) 275-280.
MDDF = measured depth from drill floor 11. Sauer, C.W.: “Mud Displacement During Cementing: A
TVDDF = true vertical depth from drill floor State of the Art,” JPT (Sept. 1987) 1091-1101.
TOC = top of cement 12. Sauer, C.W. and Landrum, W.R.: “Cementing – A
Pv = plastic viscosity (centipoise) Systematic Approach,” JPT (Dec. 1982) 2843-2852.
Ty = Bingham yield point (lbf/100ft2) 13. Harris, K., Grayson, G. and Langlinais, J.: “Obtaining
ppg = pounds per gallon Successful Shoe Tests in the Gulf of Mexico: Critical
Cementing Factors”, paper SPE 71388, 2001.
IADC/SPE 87211 9

14. Clark, C.R. and Carter, L.G.: “Mud Displacement with be provided to cover the mold and yet permit transmission of
Cement Slurries,” paper SPE 4090, 1973. pressure between the pressurizing medium (water) and the
15. Brice, J.W. and Holmes, R.C.: “Engineered Casing slurry. A height of 203mm was chosen as the maximum height
Cementing Programs Using Turbulent Flow Techniques,” which would fit conveniently in HT/HP curing cells. If it is
JPT (1964) 503-508. preferred to use a long fluid loss cell to cure the sample then
16. McLean, R.H., Manry, C.W. and Whitaker, W.W.: the tube height should be 190.5mm (7-30"). In any case, the
“Displacement Mechanics in Primary Cementing,” paper dimensions of the tube should be clearly stated on any reports.
SPE 1488, 1966.
17. Silva, M.G.P., Martins, A.L., Barbosa, B.C. and Garcia Jr, Method
H.: “Designing Fluid Velocity Profiles for Optimal The inside of the tube and all joints should be lightly greased.
Primary Cementing”, SPE 36136, 1998. After assembly it should be tested to ensure that it is leak-tight
18. Beirute, R.M. and Flumerfelt, R.W.: “Mechanics of The by filling it with water. It should then be emptied and dried
Displacement Process of Drilling Muds by Cement without disturbing the joints, either by wiping the inside gently
Slurries Using An Accurate Rheological Model,” paper with paper tissue or by blowing with air. A suitable
SPE 6801, 1977. antifoamer/defoamer must be added to the slurry. The slurry
19. Haut, R.C. and Crook, R.J.: “Primary Cementing: The should be prepared according to API RP 10B, Section 5. The
Mud Displacement Process,” paper SPE 8253, 1979. slurry should be conditioned either 1) on a pressurized
20. Bannister, C.E. and Benge, O.G.: “Pipe Flow Rheometry: consistometer to the appropriate BHCT and pressure as for the
Rheological Analysis of A Turbulent Flow System Used thickening time test (Do not exceed 85°C if the BHCT is
for Cement Placement”, SPE 10216, 1981. greater than this temperature) or 2) on an atmospheric
21. Baret, J.F.: “Why Cement Fluid-Loss Additives are consistometer to the appropriate BHCT at the appropriate
Necessary,” paper SPE 17630, 1988. heat-up rate. Meanwhile, the settling tube should be heated to
22. Sanders, T., Gusevik, R., Nida, R. and Griffith, J.: BHCT in the curing chamber or the long fluid loss cell (Do not
“Practices for Providing Zonal Isolation in Conjunction exceed 85°C if BHCT is greater than this temperature).
with Expandable Casing Jobs – Case Histories,” paper
presented at 2003 High-Tech Wells Conference and Once at BHCT, remove the slurry cup and the settling tube
Exhibition, Galveston, TX, Feb. 11-13. from the respective machines. Check that the inside of the
23. Mueller, D.T.: “Cementing of Expandable Tubulars – settling tube is dry before proceeding. Open the slurry cup and
Recent Experience in the Gulf Coast”, Presentation, BJ discard any oil that may be inside the cup. Pour the slurry into
Services, Houston, TX, (2003). the tube until it is approximately 20mm from the top. Tap the
24. Rae, P.J., Johnston, N. and DiLullo, G.: “Storable Liquid tube to dislodge any air bubbles. Fill the tube completely and
Systems for Use in Cementing Oil and Gas Wells,” U.S. tap the tube again to remove air. Verify the tube is completely
Patent No. 6,173,778 (2001). filled and put the (lightly greased) lid in place. Wipe off
25. Goodwin, K.J. and Crook, R.J.: “Cement Sheath Stress surplus cement from the outside of the tube. Carefully place
Failure”, paper SPE 20453 presented at 1990 SPE Annual the tube in a curing chamber (or long fluid loss cell) heated to
Technical Conference, New Orleans, Sept. 23-26. BHCT (or to 85°C if BHCT greater than this). Avoid spilling
26. Thiercelin, M.J., Dargaud, B., Baret, J.F. and Rodriguez, any cement. Apply 3000 psi (or 1000 psi if using the long
W.J.: “Cement Design Based on Cement Mechanical fluid loss cell). If the BHCT is greater than 83°C then heat the
Response,” paper SPE 38598, 1997. curing cell to BHCT at 2°C per minute until at BHCT.
27. Fleckenstein, W.W., Eustes, A.W. and Miller, M.G.: Maintain BHCT until the end of the test. Duration is normally
“Burst-Induced Stresses in Cemented Wellbores,” paper overnight, approximately 16 hours. Cool the cell to below
SPE 72059, 2001. 90°C and release the pressure slowly. Place the tube in water
28. BJ Cementing Engineering Support Manual, 2001. at room temperature for approximately 15 minutes. Do not
29. Recommended Practice for Testing Well Cements: API remove the tube from water until the tube is at
Recommended Practice 10B, 22nd edition, American ambient temperature.
Petroleum Institute, Washington D.C. (1997) 94- 95.
Procedure
Appendix – BP Settling Test28 • Measuring Settlement - Remove the lid and measure the
This test measures the ability of a slurry to remain a stable depth from the top of the tube to the top of hard cement.
suspension under downhole conditions. The normally-used Do this with a metal rod of similar dimensions to a
API operating free water test frequently provides misleading consistometer thermocouple.
results because of the conditions that the cement slurry is • Record this figure (in mm) as the "settlement".
subjected to (during the free water test the slurry temperature • Record the nature of the "free fluid" which is the fluid on
usually falls well below BHCT) and also because the quantity top of the hard cement in the tube (this could be free
of "free water" is only a true measurement of stability for a water, soft sludge of cement fines etc.).
limited number of slurries. • Measuring density of the column. This test should
The (reusable) tube is a cylinder with the axis vertical. Internal proceed without hesitation so that it is completed before
height is 203mm (8.00") and the internal diameter is the cement dries out significantly.
25mm.The tube is designed to be opened along its length to
facilitate removal of the hardened core of cement. A lid should
10 IADC/SPE 87211

• Remove the tube from the water, split the tube, remove
the hardened cement core and wipe off any excess grease.
• Make marks along the core at approx. 30mm intervals.
Break the core at these marks. This may be done by
holding it against a rigid metal edge and striking the core
opposite this contact point with a hammer. Keep the
sections in proper order and number them if necessary.
• Use a top-pan balance capable of measuring to an
accuracy of 0.01 gram and record the weight of each
section quickly (W1).
• Take a 250-400 ml beaker, 2/3 full of distilled water at
ambient temperature, and tare it on the balance.
• Suspend each section of core, in turn, from a noose of thin
monofilament wire (nylon or steel) and lower it into the
beaker so that the core section is completely immersed in
the water. Take care the keep the suspended section away
from the sides of the beaker and ensure that no air bubbles
are trapped.
• Record the weight on the balance (W2, weight of the
section in water), quickly.
• Density of core section, SC = weight of section in air
(WI) weight of section in water (W2)

Density of Core Section, SC = Weight of Section in Air (W1)


Weight of Section in Water (W2)
Reporting
The results should be reported like this example:
Slurry:
Class G 1.0 gps FL-33, 0.1% CD-32, 3.80 gps fresh water
Density : 15.8 ppg
Test Temperature : 145°F
Test Pressure : 3000 psi
Settlement tube length : 203 mm
Depth to firm Cement (settlement) : 3.0 mm
Density of column, top-bottom :
1.90,1.91,1.90,1.92,1.92,1.93 SG
Nature of free fluid : Water

Quantifying Settlement
After the cement has set hard in the settling tube the most
important measurement to make is the amount of settlement -
(the depth from the top of the tube to the top of the firm
cement). The cement column is also cut into sections and the
density of each section is measured. This density measurement
provides additional information, however, the amount of
settlement is the value used to measure the relative stability of
the cement slurry.
IADC/SPE 87211 11

VLA
103º20'E MA 103º40'E 104º00'E 104º20'E

CPOC
N

07º20'N CTOC NW BUNGA PM 3


BUNGA
PAKMA ORKID
BERGADING BUNGA
DAHLIA BUNGA
PAKMA BUNGA
MATAHARI
BUNGA
BERGADING DEEP-1
KEKWA
BERGADING-1 BUNGA MAWAR
CPOC THA
IL
MA AND BUNGA TERATAI PM3 CAA
LAY BERGADING H4
SIA
07º00'N PM 302 BUNGA
RAYA
BUNGA
ANGSANA-A

BUNGA KEMBOJA
BUNGA
MELOR

LAWIT- BINTANG
BUNGA
ANGSANA-D
PM 5

06º40'N JERNEH

Fig. 1 – Bergading Field Location Map


12 IADC/SPE 87211

794970 N

ENVI. DEPO
STRATIGRAPHY
TOPS HC TARGET/ LITHOLOGY DESCRIPTION EST. PORE
LITHO. PRESSURE
(m ss) TYPE DEPTH AND REMARKS
AGE GROUP
COL. (ppg)
(m ss)

Seabed
@ 57 m Soft claystone, low
PLIOCENE

COL - HIN
RECENT

saturation gas sand

6.3 ppg
TO

A/B
stringers and
occasionally coal beds.

1000
D1 1105

CO F- COL
MIDDLE - LATE

D3 1180 Alternating claystone and


MIOCENE

9 ppg
GAS
sandstone . Sandstone
D are gas bearing
reservoirs.
LCP - COS
EARLY-MIDDLE

10 ppg
MIOCENE

POSS. Predominantly
GAS

F
claystones, minor
2000 sandstones and coal
beds. Possible gas
H1 2163 bearing reservoirs. CO2
COL - HIN

H H2B GAS / COND.


2324 may be present

15 ppg
H4A 2502

H5A 2590
MIOCENE
EARLY

I2 2730

I7 2861 Claystones, sandstones


I GAS / COND.
3000 and thin coals. Gas
bearing sandstone
16 ppg
LCP

reservoirs. CO2 and H2S


gases are present.

PTD @ 2950 TVDSS


(2950 TVDSS +/- 100 m)

LEGEND :
COAL SHALE
SANDSTONE SILTSTONE

Fig. 2 – Well Bergading Deep-3 Stratigraphic Column


IADC/SPE 87211 13

Fig. 3 – Well Bergading Deep-3 Casing Design Scheme

Vous aimerez peut-être aussi