You are on page 1of 10

WELL INTERVENTION AND CONTROL

OVERVIEW
In the last 25 years, we have witnessed the introduction of tech- Well Intervention
nology that is nothing short of astonishing—things that most
of us would never have imagined during our early days in the and Control
field and on the rigs. While the biggest effect clearly has been on additional reading
overall drilling efficiency—especially real-time information while available at the
drilling—it is interesting to look at how technology has affected SPE eLibrary: www.spe.org
the well-control aspect of drilling.
SPE 107234
Today, we have the ability to look at a computer screen that provides a virtually
“First Subrigfloor HWO
instantaneous reading of numerous downhole parameters (e.g., annular pressure
Intervention System for a
near the bit, equivalent circulating density, temperature, torque, vibration, and
Floating Vessel” by Tim
lithology). Advances in surface monitoring have lead to significant improvements
Pollock, BP plc, et al.
in computerized displays of pressures, rates, and volumes, all of which are crucial
to maintaining hydrostatic control of the well.
SPE 108408
Some may wonder how wells were ever drilled without these high-tech tools. “Fishing 10,000 ft of CT
In December 1982, there were 4,530 rigs running in the US vs. approximately From a Deep Gas Well in
1,800 rigs running in the US today. When those 4,530 rigs were drilling back in Saudi Arabia” by Hasan H.
1982, measurement while drilling (MWD) was in its infancy—Schlumberger did Al Jubran, Saudi Aramco
not complete their first commercial MWD project until some time in 1980. That
means that there were more than 4,500 rigs drilling away, mostly without the available at the
benefit of any real-time bottomhole information except for what could be derived OTC Library:
from experience-based surface observations. www.otcnet.org

Anyone remember what a computer looked like in 1982? That was the year Intel OTC 18614
introduced its 80286 processor with a blazing 6-MHz speed—the computer used “Using BOPs at Pressures in
to produce this article runs 6,000 times faster. Obviously, there were no computer Excess of the Rated Working
screens in the drilling supervisor’s office giving 24-hour graphical and numerical Pressure—A Solution for
readouts in real time. If there was a computer on site, it was probably busy with High-Pressure Wells?” by
a game of pong, or maybe a basic spreadsheet. M.E. Montgomery, SPE,
WEST Engineering Services,
Technology has allowed our industry to advance to a point where pressure is et al.
manipulated for the purposes of increasing drilling efficiency and reducing for-
mation damage. It used to be that any kind of pressure at the surface was a sure
sign of a problem. Primary, well control inherently meant the absence of surface
pressure. Now we have underbalanced drilling and managed-pressure drilling
that seem completely foreign to this long-held concept.

Numerous articles could be (and probably have been) written about the risk/reward
aspects of where technology has allowed us to venture. Sometimes it is just interest-
ing to sit back and look at where we were not that long ago. You cannot help but
wonder where continued advancements will lead over the next 25 years. JPT

David Barnett, SPE, is vice president, Engineering Services for Wild Well Control,
a company specializing in firefighting, well-control, and engineering services. He
has more than 20 years of drilling, snubbing, coiled-tubing, and well-control experi-
ence. Barnett’s experience includes work as a drilling engineer, rig superintendent,
snubbing supervisor, and well-control specialist. He has been involved with the
planning and implementation of numerous relief-well, high-pressure-snubbing, and
well-recovery operations. Barnett was a team leader in the Kuwait project and has
worked on well-control projects in many countries around the world. He holds a BS
degree in mechanical engineering from the University of Houston and serves on the
JPT Editorial Committee.

66 JPT • JANUARY 2008


WELL INTERVENTION AND CONTROL

Wellbore-Strengthening Technique for Drilling Operations

A significant increase in formation frac-


ture resistance can be achieved as a
result of a fracture sealing or plug-
ging mechanism induced by a particle
“screenout” effect resulting from the
drilling fluid being loaded with an
adequate amount of narrowly sized
granular materials. Such an increase in
formation fracture resistance is particu-
larly valuable in helping to drill through
depleted zones without losing fluid,
strengthening the weaker formations
that usually require additional casing
strings for protection, avoiding lost cir- Fig. 1—Well C WS treatment design; TVD=true vertical depth,
culation during cementing operations, MD=measured depth, and MW=mud weight.
and drilling high-angle well sections
with high mud weights that normally material, heating the wellbore, and use down pressure, not the far-field frac-
would not be possible because of low of rigid-plug-forming pills. The full- ture gradient.
formation fracture gradients. length paper focuses on a WS method
that uses mud with special bridging Case Histories
Introduction and fluid-loss formulations. The full-length paper presents four
Lost circulation is perhaps the most In the mid-1980s, a joint-industry case histories that provide a sampling
costly mud-related drilling problem. project (JIP) was conducted to deter- of the various ways in which the WS
Not only is rig time lost, but large mine why oil-based mud apparently technique can be applied. Although
volumes of expensive drilling fluids caused a lower fracture gradient than the optimal way to apply this tech-
are lost to the formation. Lost circu- water-based mud (WBM). Data from nique is to drill the weak zone with the
lation also can lead to severe well- this JIP were analyzed, and it was granular material carried in the entire
control incidents. In response, several determined that they did not fit ordi- mud system (the drill-ahead method),
technologies referred to as “wellbore nary fracturing models of the time. three of the four examples are actually
strengthening” (WS) are being devel- These researchers noted that certain cases of the formation-integrity-test
oped to prevent or correct lost circula- lost-circulation materials (LCMs) actu- (FIT) -type method.
tion. WS means that the breakdown ally seemed to increase the pressure
pressure of the wellbore is greater after required for fracture propagation to Case A. Wells in one field had lost
the treatment than before. These meth- more than what was seen with untreat- circulation zones in two sandstone
ods include use of special granular ed drilling mud. They conducted labo- formations because neither formation
ratory tests outside the JIP to verify could withstand mud densities greater
This article, written by Assistant their observations. The result was the than 9.0 lbm/gal. The limestone reser-
Technology Editor Karen Bybee, con- development of a loss-prevention mate- voir below them is pressured at 10.1
tains highlights of paper SPE 105809, rial (LPM) that was found to inhibit the to 10.5 lbm/gal. The traditional solu-
“Further Development, Field Testing, extension of a fracture tip. tion had been to set an 8,050-ft string
and Application of the Wellbore- LPM was tested in the field several of 7-in. intermediate casing, drill the
Strengthening Technique for Drilling times and was quite successful. It was remaining 650 ft of wellbore, and then
Operations,” by Giin-Fa Fuh, SPE, and able to increase the wellbore break- hang a liner from the intermediate cas-
Dave Beardmore, SPE, ConocoPhillips down pressure in several cases, and ing across the pressured limestone res-
Co., and Nobuo Morita, SPE, Waseda subsequent application of the technol- ervoir. This approach has two distinct
U., prepared for the 2007 SPE/IADC ogy has allowed a casing string to be disadvantages. First, if the well turns
Drilling Conference, Amsterdam, 20–22 skipped in several cases. The technique out to be a dry hole when total depth
February. affects only the near-wellbore break- (TD) is reached, an expensive 8,000-ft

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.

JPT • JANUARY 2008 67


string of casing has already been run For the 2,250 ft of hole that was ing was greater than that estimated to
and cemented in the wellbore. Second, drilled with 30/60-mesh LPM in the be applied in the annular space during
the casing/liner combination is more mud, 60-mesh screens were used on the cementing when the lost-circulation
expensive than a single string of pro- shakers to provide some mud cleaning problem occurred.
duction casing. for that long interval. Some of the LPM Several significant changes were
The first attempt was to eliminate that was removed by the shakers was made to the well plan before drilling
the 8,050-ft intermediate-casing string recycled into the mud, and significant the next well. Because lost circulation
while drilling Well A1. The decision additions of new LPM were made to was prevented while drilling Well A2,
was made to leave the weak-sandstone replace what was not recovered from a more aggressive plan was used to
intervals exposed to the 10.0-lbm/gal the shakers. This was operationally drill Well A3. A 77/8-in. bit was used
mud used to drill the pay. In doing so, very difficult for such a long section. to drill a smaller production hole out
it was assumed that the occurrence of To avoid losing circulation during of the smaller 85/8-in. intermediate-cas-
lost circulation was inconsistent within the cement job, plans were to perform ing string set at 3,500 ft. Fresh water
the weak zones in nearby fields and that a two-stage job—placing a 15.6-lbm/gal was used to drill through the weak
lost circulation in this field might not cement from TD up to 8,000 ft dur- formations, and the hole was displaced
be a problem. Lost circulation occurred ing the first stage and then circulat- above the top of the pay at 8,500 ft with
into the uppermost weak sandstone, ing a low-density foam cement from the same type of LPM drilling fluid
and the resulting reduction in hydro- 8,000 ft up into the surface casing. The used to drill Well A2. Because of the
static head allowed an influx of oil and foam cement would limit hydrostatic higher formation pressure at this loca-
gas from the limestone producing for- head, preventing lost circulation into tion, the mud weight was increased to
mation. Although the well control was the exposed weak formations below 10.5 lbm/gal. The shakers were bypassed
skillfully managed without incident, 6,500 ft. After completing the first stage, for this short 200-ft interval to allow the
the hazardous conditions necessitated the differential valve tool at 8,000 ft LPM to remain in the mud system. No
finding a lower-risk solution, and LPM was opened and approximately 15 bbl losses were detected at any time during
technology for WS was presented as a of mud was pumped before circulation drilling or casing operations, and the
potential solution. was stopped to switch from service- well was completed successfully and
Because of apprehension about the company pumps to the rig pumps (i.e., put on production. In this field case,
effectiveness of the LPM concept, a plans were to circulate for approxi- a minimum of 1.5 lbm/gal in WS was
modified approach was pursued in mately 4 hours before beginning the achieved and use of LPM was proved
Well A2. An 83/4-in. hole was drilled second-stage foam-cement job). When to be effective in inhibiting fracture ini-
out from under the 95/8-in. interme- attempts were made to reinitiate cir- tiation and propagation in the depleted
diate-casing string at 3,500 ft. Fresh culation, complete losses occurred as formations. A 15 to 20% cost savings
water was used to drill to 6,450 ft, at the result of a pressure surge, and on this well was attributed to the use
which point the wellbore was displaced circulation was never recovered. It was of LPM technology. Modification of
with a 9.2-lbm/gal LPM mud consisting estimated that at the time, the spacer the well plan to drill smaller holes also
of salt gel, starch, and LPM. The LPM and some of the cement from the first resulted in higher penetration rates,
was 40 lbm/bbl of 30/60-mesh walnut stage were directly opposite the weak reduced drilling-fluid volumes, less cut-
hulls. As a precaution, the decision formations. The pressure surge from tings waste, and improved wellbore sta-
was made to drill into the top of the breaking circulation caused a fracture bility. Another significant change made
weak zones with weighted LPM fluid to develop, and because the spacer and for the Well A3 application was to
and determine the effectiveness of the first-stage cement slurry did not contain include LPM in the cement formulation
LPM technology without exposing the LPM, there was no way to prevent the at the same concentration used in the
several hundred feet of the entire weak fracture from propagating. Fortunately, drilling fluid. This change prevented
interval to an excessive mud weight. subsequent logging indicated top of lost circulation during the two-stage
If lost circulation occurred at the top cement at approximately 6,640 ft (near cement job, yielding a competent col-
of the weak zones and the LPM was the top of the weak sandstones), so the umn of cement up into the surface
ineffective, the test would be discon- producing zone was protected and iso- casing. After using LPM successfully on
tinued, and the mud weight could lated sufficiently. Wells A2 and A3, the technology was
be reduced to less than 9.0 lbm/gal. It would have been preferable to used in subsequent wells drilled in the
The intermediate hole then would be avoid the lost-circulation problems same area. In each of the wells, the LPM
drilled to 8,050 ft, and the 7-in. inter- experienced during the cementing technology was used to eliminate the
mediate-casing string would be set as operations. However, the lost-circu- intermediate-casing string and to slim
usual. However, no losses occurred lation problems clearly proved that down the diameter of the wellbore.
during drilling operations when the including LPM in the drilling fluids did
mud weight was increased from 9.2 prevent lost circulation from occur- Case C. Well C had a weak zone at
to 10.1 lbm/gal with the addition of ring when drilling through the weak TD with high mud weight required for
drilled solids and salt. The well was formations. Complete circulation loss shallower-zone stability. Well C is one
drilled successfully to TD at 8,700 ft, during the cementing phase indicated of many directional wells with a hori-
and 51/2-in. production casing was run that lost circulation also would have zontal production section drilled from a
and cemented. As a result, a net WS occurred during the drilling phase central location in an existing field. The
of at least 1.1 lbm/gal was achieved in because the hydrostatic head caused by 81/2-in. intermediate section is drilled
this case. the 10.2-lbm/gal mud used while drill- with a WBM at a sail angle of approxi-

68 JPT • JANUARY 2008


mately 65° and into the pay with a final cement job because of the regulatory normal 11.8- to 13.0-lbm/gal fracture
inclination of 90°. Then, 7-in. casing is cementing requirements. gradient for the pay sand. The treatment
run and cemented at TD. The produc- The pay-sand fracture gradient varies was considered a success because there
tive sandstone reservoir is depleted in with local pore pressure. The reservoir were no losses, and a remedial cement
some areas, but with the usual 9.8- to has linedrive producers and injectors. job was avoided.
10.0-lbm/gal mud in the intermediate Depending on the nature of the perme-
hole section, losses typically do not ability of the reservoir and its connec- Discussion
occur during drilling. While running tion with either injectors or producers The mechanism of increased fracture
the 7-in. casing to bottom, mud is rou- at the point of penetration, the pay sand resistance or WS depends highly on
tinely circulated at three depths to aid could have a higher (or lower) than the LPM concentration used in relation
in getting the 7-in. casing to bottom. normal fracture gradient. The fracture to the degree of formation permeabil-
Occasionally, losses occur during run- gradient of the pay at virgin pore pres- ity. Numerical modeling is in progress
ning and cementing the casing. sure is approximately 13.0 lbm/gal. The to analyze the effect of LPM and its
While drilling Well C, shale caving 7-in.-casing shoes typically undergo concentration on fracture propagation
began to occur in a shale zone above a 12.5-lbm/gal FIT before drilling the during the dehydration of LPM and
the pay sand. Although the intermedi- horizontal section. Approximately one- mud solids. The modeling indicates
ate-section TD was reached, the caving fourth of the wells produce leakoff that at a low LPM concentration (i.e.,
hole section was abandoned and had to before reaching 12.5-lbm/gal EMW. 5 to 10 vol%), the LPM dehydrates
be redrilled. The shale caving was con- The intermediate section was and forms a sealing barrier around the
trolled by increasing the mud weight drilled to TD (15,844 ft) with 9.7- to fracture tip (in permeable formations)
to 10.0 to 10.6 lbm/gal. On the basis 9.8-lbm/gal mud. Caving shale and that significantly increases the apparent
of experience and annular-pressure- packoffs became a serious problem. fracture toughness to prevent fracture
drop modeling, losses were not expect- Mud weight was increased, and some propagation. For a medium LPM con-
ed while drilling to TD. Equivalent breathing (loss of mud followed by centration (i.e., 10 to 30 vol%), the
circulating densities (ECDs) would return of part of the lost volume when LPM dehydrates at the mouth of the
not exceed 11.9-lbm/gal equivalent pumps are shut off) was noted at a induced fracture for permeable and
mud weight (EMW). But losses were 10.6-lbm/gal mud weight. However, it less-permeable formations, and acts
expected while running and cementing is believed the losses were aggravated like a sealant and increases the rupture
the 7-in. casing, when ECDs would by numerous packoff instances. This strength of the borehole. At a much
approach 13.6-lbm/gal EMW. original hole had such bad caving shale higher LPM concentration (i.e., 30 to
After discussing the situation, the that a cement plug was set over the 50 vol%), the LPM would plug the
team developed the idea of spotting bottom 2,000 ft. The hole was redrilled mouth of the induced fracture even
a mud pill containing an effective with 10.0- to 10.6-lbm/gal mud. At TD for formations with zero permeability
amount of LPM on bottom across the of the section (15,758 ft), the WS pill because of geometrical factors
pay sand before pulling out of hole was spotted on bottom to cover the pay All four of these case histories in
with the drillpipe before running the sand fully, with some additional volume the full-length paper were conducted
casing (Fig. 1). The concept was that in case some minor losses occurred. in what would be described as “hard
when the high ECD was applied to the The drillpipe was pulled above the pill, rock.” None were conducted in uncon-
hole during circulation of the casing and 11.0-lbm/gal mud was circulated solidated sands, silts, or clays.
on the way to TD, the WS material and conditioned above the pill with no The drill-ahead type of application
would inhibit fracture growth, prevent losses. The pressure-while-drilling tool should be the most effective because
major losses, and possibly strengthen in the hole recorded a maximum sus- the LPM always is available to inhibit
the wellbore before the cement job. In tained ECD of 11.8-lbm/gal EMW while fracture growth. However, long hole
this case, getting cement 2,300 ft above circulating, with spikes to 12.1-lbm/gal sections may preclude its use because
the shoe was a regulatory requirement EMW during packoffs. of solids buildup in the mud. Therefore,
for this waterflood injection well. The 7-in. casing was run in the hole, development and understanding of the
Formulation of the WS pill to be with circulation conducted at 3,453 ft FIT-type method is warranted.
spotted on bottom was similar to what (surface shoe), 6,945 ft, 10,420 ft, and One of the most important aspects
had been successful before. The plan 13,828 ft at 3 to 4 bbl/min. The casing of either of these methods that use the
was that the pill would stay on bottom was reciprocated during these stops. special granular material is that they are
while the drillpipe was pulled out of the The casing was worked through tight preventive treatments. Theoretically,
hole and while the 7-in. casing was run hole at 14,965 to 15,640 ft. No losses the granular material will work best
to bottom. The pill would be circulated occurred while running or circulating only on newly forming fractures. It is
out of the well when the 7-in. casing casing. At TD, the casing was circulated conceivable that the suitable concen-
got to bottom and would be circulated at 4 bbl/min, and the mud weight was trations of granular materials might be
before cementing. reduced from 11.0 to 10.6 lbm/gal. The more effective than other LCM mixes in
The main risk of this WS application casing was cemented with a 12.5-lbm/gal situations of induced- or natural-frac-
was that if the pill failed to strengthen spacer and a 15.8-lbm/gal cement ture lost circulation that have already
the pay sand, it was likely that there slurry at 4.5 bbl/min with no losses. occurred, but theoretically, and so far
would be lost circulation while run- The cement simulator indicated that in practice, they have not been able to
ning the casing or during the cement the maximum ECD was approximately strengthen formations that have already
job. This would necessitate a remedial 13.6 lbm/gal, which is well above the suffered large losses. JPT

70 JPT • JANUARY 2008


WELL INTERVENTION AND CONTROL

Wellbore Strengthening in Shale

The full-length paper demonstrates that One of the most popular wellbore- will prevent further seepage of fluid
wellbore strengthening in shale is fea- strengthening technologies is induc- through a leaky seal on the bridge, will
sible. A treatment pill was developed ing an increase in wellbore stresses by prevent flowback toward the wellbore
in the laboratory and field tested suc- use of sized-particulate additions to the when hydrostatic pressure is reduced
cessfully at a US land-based location. drilling fluid. The common name for temporarily, and will help bear the
The treatment consisted of a blend of this approach is “stress caging.” The stress across the fracture. This solution
particulates (known as stress-cage sol- principle is to increase the hoop stress can be enhanced further if the mate-
ids) and proprietary crosslinked gelling around the wellbore by using fractures rial also shows a propensity to adhere
polymers that set with time. Properties to cause stress changes in the rock. to shale. In this manner, the propping
of the system such as compressive These fractures would be held open particulate can be introduced to the
strength, adhesion to shale, and sensi- with bridging material, thereby creating fracture with the proper size and con-
tivity to temperature and pressure were the stress-cage effect, or strengthening centration, allowing a desired set while
evaluated. Modeling work was con- of the wellbore. This bridge of particles avoiding undesired pressure transmis-
ducted to engineer the size and con- across the mouth of the fracture must sion and then adhering to the shale
centration of bridging solids required. have low permeability to provide pres- fracture, thus providing a permanent
sure isolation of the fluid in the well- bridge at the opening of the fracture.
Introduction bore from that in the fracture, thus pre-
Drilling in depleted zones, or inter- venting any further fracture elongation. Field Trials
vals where high-pressured formations The degree of stress caging relative to BP has conducted several stress-cage
are interbedded with normally and the native stresses is a function of both trials in shale. The previous trials car-
abnormally pressured layers, has cre- fracture width and radius and the rock ried out in the US Arkoma basin were
ated the need for these strengthening properties (i.e., Young’s modulus and somewhat encouraging. Pressures
technologies. The goal is to increase Poisson’s ratio). To achieve this effect, greater than 5 lbm/gal above the forma-
the fracture resistance of weaker for- the proper type, size, and amount of tion fracture pressure were achieved as
mations and thereby avoid mud losses, loss-prevention material (LPM) must be long as the stress-cage pill remained in
wellbore instabilities, and the poten- used. Using the stress-cage concept and the hole. The pill was formulated in oil-
tial loss of the drilled interval. The associated factors, drilling fluids can be based mud (OBM) and contained sized
primary consequences of these unde- designed to maximize this strengthen- marble and graphitic bridging solids,
sired events are an increase in nonpro- ing effect and minimize losses and other but was of a nonsetting type.
ductive time and associated costs. In problems encountered when drilling When the pill was displaced after the
addition to avoiding these problems, significantly above the fracture gradient. squeeze, and regular mud was circu-
the economic benefits that wellbore In fact, use of these LPMs for wellbore lated, the strengthening effect was lost.
strengthening can provide are the pos- strengthening has led to several comple- It appeared the bridging particles were
sible elimination of a casing string or, mentary technologies. not held sufficiently strongly within
more importantly, the ability to reach Although the stress-cage approach the fracture and subsequently were
deeper reservoir targets. is used widely, the application primar- stripped out.
ily has been confined to permeable
This article, written by Assistant Technology formations, even though strengthening Gel System. A new trial was conducted
Editor Karen Bybee, contains highlights impermeable rocks (e.g., shale) also in the Arkoma basin to test the novel
of paper SPE 110713, “A New Treatment often is required. The key to stress gel system, with the goal of achieving
for Wellbore Strengthening in Shale,” by caging shale would be to design a tech- a more permanent strengthening effect.
Mark S. Aston, SPE, Mark W. Alberty, nology (i.e., chemistry, technique, or The test was conducted in a vertical well,
SPE, and Simon Duncum, BP plc, and both) whereby once the sized bridging across 50 ft of Atoka shale in an 83/4-in.
James E. Friedheim, SPE, and Mark particulates prop open the fracture, hole at approximately 4,020 ft. The
W. Sanders, SPE, M-I Swaco, pre- they cannot be removed easily. One shale was directly below the 95/8-in.-cas-
pared for the 2007 SPE Annual Technical approach is to transport the bridging ing shoe. The Atoka formation is a fairly
Conference and Exhibition, Anaheim, particulate in a settable medium that unreactive brittle shale. An OBM with a
California, 11–14 November. solidifies the fracture. The solidification 9.3-lbm/gal density was used.

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.

JPT • JANUARY 2008 71


Baseline Data. A 1,928-psi forma- held during the squeeze, but after drill- to the original design value for the pill.
tion-breakdown pressure was mea- ing out the set treatment and circulating The drilling team preferred to leave the
sured, and upon shut-in, the pres- clean, there was only a limited strength formation in a strengthened condition,
sure was bled back to approximately increase. The fracture opened again at and so it was not possible to increase
1,500 psi. As expected, the bleed- 1,800 psi. It seemed that insufficient the pressure further to determine the
back value coincides roughly with squeeze pressure had been applied to actual breakdown point.
the reopening pressure (1,525 psi) force the bridging-solids/gel system
seen on repressurization. The differ- fully into the mouth of the fracture. Discussion
ence between initial breakdown and The slope of the shut-in curves for
reopening of approximately 400 psi Second Gel-Pill Application. The sec- the various FITs was quite variable.
(1,928−1,525 psi) represents the ten- ond pill formulation and procedure Accordingly, obtaining a low-pressure
sile strength of the rock. were similar to those of the first pill, (1,400 psi) baseline for each FIT test
The bottomhole shut-in tempera- except a higher squeeze pressure was was desirable because this is less than
ture was logged as 107°F, which was applied. A 2,500-psi squeeze pressure the fracture-reopening pressure and
considerably lower than the predicted was used, which was approached in can be used as a reference point; doing
120 to 127°F. Rather than redesign- stages over a 2-hour period while the so helps avoid erroneous conclusions.
ing the chemistry of the gel system on treatment was still liquid. The pres- Some pressure decay on shut-in is nor-
short notice, the decision was made to sure actually climbed to approximately mal and can be caused by a number of
extend the shut-in period. Therefore, 3,000 psi toward the end of the shut-in effects, including temperature changes,
long shut-in times of approximately period, most likely as a result of an seepage of fluid into the fracture, or
20 hours were needed for the trial. increasing wellbore temperature and leaks introduced in the equipment.
the heating of surface lines (night vs. Hence, variations between different FIT
Pill Formulation and Placement. day temperatures). It is remarkable tests might be expected.
The bridging package for the pill was that the formulated pill could hold the Survival Test 3 produced a remark-
designed by running in-house soft- 2,500 psi while still fully liquid, but it ably flat shut-in pressure, and the big-
ware that predicts fracture widths from was expected that it could hold high gest contrast with this was Survival
geophysical data. A fracture-opening pressures (3,000 psi) once set. Test 4 where there was a fairly steep
width of 640 μm was predicted, which After drilling out, survival tests slope on shut-in. The slope in this final
would allow the wellbore pressure to were carried out. For each survival test does not change significantly as the
exceed the minimum horizontal stress test, mud was circulated for a set pressure is increased and, most impor-
by 500 psi. The bridging-solids blend period and the bottomhole assembly tantly, is similar to that of the reference
then was selected to cover a wide range was rotated and moved up and down test (1,400 psi). Therefore, the authors
of sizes (approximately 2 to 800 μm), through the strengthened shale to concluded that the slope does not
allowing for possible variations in frac- avoid washout. The survival was then imply failure of the stress cage and is
ture width and ensuring that a good assessed by formation-integrity-test most likely caused either by differences
seal was obtained on the fracture. A (FIT) testing. The FIT pressure was in the cooling of the well as a result of
high concentration of bridging solids taken to 1,400 psi initially to establish differences in circulation times or by a
included 25 lbm/bbl of coarse mar- a baseline, and then to 1,700 psi in leak in the system introduced between
ble, 25 lbm/bbl of graphitic solids, Tests 1 through 3 (175 psi above the the different circulation stages.
20 lbm/bbl of medium marble, and fracture-reopening pressure), and up
20 lbm/bbl of fine marble. to 2,080 psi in Test 4 (555 psi above Conclusions
Additional fine marble was added the fracture-reopening pressure). 1. Shale can be strengthened reliably
as required to increase the weight Tests 1 through 3 show that a by use of stress-cage techniques using a
of the pill to the desired value strengthening effect was indeed settable pill containing bridging solids
(10.5 lbm/gal). Laboratory work had achieved, and that the treatment could with sufficient compressive strength
shown this to be beneficial, and it survive increasing periods of mud cir- and shale-adhering qualities.
would increase the compressive culation. The total circulation time at 2. Proper design and field implemen-
strength of the set material. the end of Test 3 had been 2.75 hours tation of this technology are essential
A balanced-plug technique was used. at half the normal circulation rate— to creating stress cages successfully
The procedure was to pump a 10-bbl 220gal/min. Even more impressive in shale.
diesel spacer, followed by 16 bbl of is Survival Test 4. In this final test, 3. Wellbore strengthening through
treatment, and then 2 bbl diesel so the circulation rate was increased to stress caging can be accomplished
that column heights/densities were bal- 440 gal/min for a further 2 hours, and a without the dependence upon fracture
anced in the annulus and drillpipe after final series of FIT tests was performed. collapse caused by leakoff, by maintain-
displacement. Foam wiper balls were The surface pressure was increased ing a bridge at the mouth of the fracture
placed on either side of the treatment. in stages and reached an impressive and preventing additional fluid leakage
2,080 psi before the test was termi- into the fracture.
First Gel-Pill Application. In the nated. This is approximately 150 psi 4. The demonstration of the feasibil-
first attempt, the pill was squeezed for greater than the original formation- ity of the application of stress cages to
19 hours at 2,000-psi surface pressure breakdown pressure (1,928 psi), and shale opens the door for further devel-
(72 psi above the initial formation- 550 psi more than the fracture-reopen- opment and application of stress cages
breakdown pressure). The pressure ing pressure. This value is in fact close in shale. JPT

72 JPT • JANUARY 2008


WELL INTERVENTION AND CONTROL

Hydraulic Blowout-Control Requirements

The full-length paper presents equa- 250


tions and guidelines for determining
requirements for pump rates, number 200
Pump Rate, bbl/min
of wells, and kill fluids required to kill
a blowout. A comparison with selected
150 X>1
field cases demonstrates that the for-
mulated guidelines match the require- X = 0.75
ments to actually kill these blowouts. X = 0.5
100
This comparison gives confidence that X = 0.25
the guidelines presented are a good
starting point for contingency plan- 50
ning. Several blowout scenarios for a
giant North Sea gas field are exam-
ined to demonstrate application of 0
the guidelines. 4 5 6 7 8 9 10 11
Conduit ID, in.
Introduction
Because of advances in drilling and Fig. 1—Required pump rates vs. blowout-conduit inside diameter (ID).
production technology, blowouts have
become an increasingly rare phenom- basis of the emergency-response plan, outflow, is inaccessible, or is damaged
enon. A report published by the United preparations can be made to take action to the extent that repair is impossible.
Nations Environment Program that cites rapidly after an incident is reported. Such Examples of these situations include
statistics from the US Gulf of Mexico preparations would include ensuring the the following.
(GOM) and the North Sea shows the availability of a nearby rig for drilling a • Underground blowouts, where the
probability of shallow-gas blowouts in relief well, sufficient tubular-goods stock flow of hydrocarbons enters a shallow
exploration wells to be approximately to complete a relief well, and pump low-pressure zone (e.g., through a cas-
one in every 200 wells; for workover capacity to kill the blowing well. With ing leak). In the case of a shallow leak,
operations, one in every 2,500 wells for these elements in place, relief efforts can the fluids may reach the surface some
oil wells and one in every 1,000 wells for be started quickly, thus reducing the distance from the well.
gas wells; and for infill wells, less than one time to regain control of the well. The • Subsea wells.
in every 10,000 wells drilled. Because of plan should consider the possible modes • Wells and platforms that have suf-
this, blowout prevention is a top priority of failure and the response to these fail- fered such damage that a safe position
in drilling and well intervention. A con- ures to restore well integrity. cannot be created for staff to work on
tingency plan must be formulated before In the majority of cases, “capping” the well. The Bay Marchand incident,
spudding development wells to limit the of the well (i.e., replacing the damaged with 11 wells on fire on a single plat-
consequences of a blowout by prescrib- wellhead) is the most efficient method form, was an example of this situation.
ing an adequate emergency response to to stop flow from the well. To illustrate, In these cases, to restore the mechan-
a possible well-control incident. On the most of the more than 780 wells that ical integrity of the well, inflow must be
were sabotaged and set on fire during inhibited by balancing the pressure of
This article, written by Assistant Technology the 1991 Gulf War in Kuwait were the producing formations with a fluid
Editor Karen Bybee, contains highlights of brought under control within 8 months column in the well or by blocking the
paper SPE 105612, “Hydraulic Blowout- by capping. With this success, well- entry of oil and/or gas into the well
Control Requirements for Big-Bore and control companies demonstrated that by saturating the near-wellbore region
HP/HT Developments: Validation With even under harsh post-war conditions, with water or brine.
Field Experience,” by P. Oudeman, capping technology could deal with The latter method, called “flooding,”
SPE, Shell Intl. E&P, prepared for the almost any form of wellhead damage. requires the injection of large amounts
2007 SPE/IADC Drilling Conference, Unfortunately, capping is of little use of water into the near-wellbore forma-
Amsterdam, 20–22 February. when the wellhead is not the point of tions through one or more relief wells

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.

JPT • JANUARY 2008 73


and is usually a cumbersome and time- GOM in 165 ft water depth. In June abandoned and the rig was evacuated
consuming process. To illustrate, in Bay 1979, control was lost when the well when gas broke around the 20-in. shoe
Marchand, this required drilling wells to reached 11,900 ft, and total losses were and bubbled up under the platform.
within 25 ft of each of the blowing wells. observed. The well blew out after trip- Eight days later the gas ignited, and in
Ten of the wells were killed this way. One ping out the drillstring, and the rig was 2 weeks, the rig and platform disap-
remaining well had to be capped after evacuated when it caught fire. This was peared into the Persian Gulf.
100,000 bbl of seawater was pumped the beginning of a blowout that lasted The gas originated from the Asmari
without any noticeable effect. Hence the for 9 months and 22 days, a period limestone formation, which had a
first process, hydraulic blowout control, over which an estimated 4.2 million 115-ft gas column at the position of
usually will be the preferred option. bbl of oil were spilled into the GOM, L-3. This formation is vuggy and frac-
For successful hydraulic blowout the second-largest oil spill ever after tured, with permeabilities of several
control, often referred to as “dynamic the deliberate release of approximately hundred md. The 171/2-in. hole was
kill,” it is essential that kill fluids be 5.7 million bbl in the Persian Gulf dur- open to flow of gas and water from
pumped downhole into the blowout ing the 1991 Gulf War. this formation. A hole this size from
well at such a rate that the well can no The well was blowing through 95/8-in. top to bottom would require a kill
longer lift oil and/or gas to surface and casing to surface. The exact pressure rate of more than 700 bbl/min and
the well ceases to flow once sufficient was unknown, but the fact that total 6 relief wells. However, the drillstring
pressure has built up at the sandface. losses were experienced at 1.12-specific was left in the hole, which gave an
This means that for contingency plan- gravity (SG) mud weight suggests that additional resistance to flow. Flow was
ning for a new well, two primary ques- the pressures were close to hydrostatic. not directly to surface but around the
tions need to be answered. According to Fig. 1, this requires a 20-in.-casing shoe and through the for-
1. What kill fluid should be pumped kill rate of approximately 140 bbl/min. mations surrounding the well. Heavy
into the well and at what rate to control Static formation pressure, assuming a kill fluids could be used to reduce the
the well flow when mechanical control hydrostatic gradient, exceeded 5,000 psi kill rate required. A more refined analy-
is not an option? [i.e., the maximum pump rate that could sis taking these factors into account
2. Will it be technically feasible to be achieved through a well with 95/8-in. indicates a kill rate of approximately
pump this rate into the well given the casing and 5-in. drillpipe, 132 bbl/min 300 bbl/min for seawater. Eventually
limitations of the kill string and pump (Fig. 2 in the full-length paper), was four relief wells were drilled, and dur-
equipment? insufficient]. At least two relief wells ing five kill attempts, fluids (including
For normally pressured, medium- were required. polymers to combat fluid losses) were
sized oil and gas wells, the feasibility Two relief wells, Ixtoc 1-A and Ixtoc pumped downhole into the well at
of hydraulic blowout control has been 1-B, were spudded. Ixtoc 1-B intersected rates as high as 230 bbl/min without
proved in a large number of reported the blowout first, and injection of sea- success. The well bridged and ceased
cases. Considerably less experience is water with rates averaging 50 bbl/min flowing approximately 7 months after
available in areas such as high-pres- started right away. Although this had a the blowout started.
sure/high-temperature wells, big-bore positive effect on the amount of oil pro- This case demonstrates that hole size
gas wells, and highly prolific wells such duced, it did not kill the well. Rates as is a major factor for blowout control.
as long horizontal wells. high as 100 bbl/min were achieved with Whereas high pressure can be coun-
Answering the primary questions the 14,000 hp available, but these could tered by high-density kill fluids, a lack
of hydraulic blowout control for these not be maintained because of the rough of resistance to flow will in all cases
types of wells, with their high blowout seas, which are common in the Bay of require high pump rates.
potential and associated potential dam- Campeche during winter. Two months
age, is of utmost importance. Here, the later, Ixtoc 1-A intersected the blowout. Arun-C-II-2. The blowout of Mobil
answers will be sought by applying the Attempts were made to kill the blowout Indonesia’s Arun C-II-2 is considered
validated knowledge obtained in killing by simultaneously pumping down the the largest gas blowout ever. The res-
“conventional” well blowouts to less- two relief wells with mud density as ervoir pressure at a 9,650 ft true verti-
conventional wells. This approach will great as 1.8 SG, but without success. cal depth (TVD) was 7,100 psi. The
indicate particular areas of concern and Unfortunately the pump rates are not temperature was 230°F. The well was
provide general guidelines for the design specified in the literature. Gradually, blowing 0.6-SG lean gas through the
of a relief well and the kill job. These can the water cut of the produced fluids annulus of a 95/8-in. casing and 5-in.
be considered a good starting point for increased. Continued pumping of sea- drillpipe. Measured depth of the well
the relief-well and kill-job design. water, possibly in combination with was 10,210 ft.
The full-length paper presents simpli- water breakthrough in the well, finally In the Arun case, the required
fied equations for the minimum pump killed the blowout on 17 March 1980. 126-bbl/min kill rate was much high-
rate required to kill overpressured wells er than the 80 bbl/min calculated or
and big-bore wells, required hydraulic Fateh L-3. The Fateh L-3 develop- the 118 bbl/min indicated by Fig. 1.
horsepower of the pump, and required ment well offshore Dubai had reached Previous authors have attributed this
pumping time. 4,180 ft when the kick occurred in the to leakoff in the very-high-permeability
171/2-in. wellbore. A 30-in. conduc- reservoir between the relief well and
Comparison With Field Cases tor had been driven to 450 ft, and a the blowout well. The leakoff resulted
Ixtoc 1. Ixtoc 1 was 58 miles north- 20-in. casing was set and cemented in only a fraction of the kill fluid actu-
west of Ciudad del Carmen in the at 1,310 ft. Kick-control efforts were ally flowing up the blowout well.

74 JPT • JANUARY 2008


The discussion of the field cases condensate and 2,500 MMscf/D gas top of the reservoir, respectively. Two
shows that the theoretical relations to be produced by 24 subsea wells in subcases were considered for Scenarios
between well parameters, such as four seabed templates. Condensate/ 4 and 5.
geometry and pressure regime, and gas ratio is estimated at approximately Of the scenarios considered, Scenario
the actual relief effort required are in 18 bbl/MMscf; water production is con- 2, blowout through 7-in. tubing during
fair agreement. This allows the final sidered negligible. The 150-ft-thick sands production, will have the highest resis-
step, quantification of the relief effort with an average permeability of 500 md tance to flow and will have the lowest
required for a proposed development. are normally pressured (4,200 psi). required pump rate. The method pre-
A subsea well at a depth of 2,788 ft sented in the full-length paper indicates
Contingency Planning below mean sea level will blow out that 70 bbl/min of seawater will be
To illustrate application of the princi- against a surface pressure of 1,270 psi, required to kill this well. A single relief
ples presented in the full-length paper, assuming seawater SG of 1.03. This well will be sufficient, and 6,000 hhp
the case of the Ormen Lange gas field, means the effective reservoir pressure will be needed.
offshore Norway, will be examined. is 2,930 psia. For a vertical distance of The worst case is Scenario 4 with
The Ormen Lange field, discovered 6,562 ft between the reservoir and the 133/8-in. casing and flow in the open
in 1997, is in the Storegga slide area. wellhead, this gives a hydrostatic ratio, hole/casing annulus with no drillpipe
Water depth is 2,789 to 3,089 ft. It is X, practically equal to one. in the hole, and relief-well intersection
the second-largest gas field in Norway. Several scenarios were examined, just below the 133/8-in. casing. This
With gas reserves of close to 13 Tcf and predictions were made by use of scenario requires 310 bbl/min. Three
and development costs of approxi- the equations and figures presented in relief wells would be required. High-
mately USD 0.5 billion, Ormen Lange the full-length paper. Scenarios 1 and density kill fluids could reduce this
ranks as the largest development in 2 were blowouts to the seabed during to two wells. Only when the reservoir
the European offshore arena. The res- production through 95/8-in. tubing and pressure drops to less than 2,470 psi,
ervoir is approximately 24 miles long 7-in. tubing, respectively. Scenario 3 reducing the hydrostatic ratio to less
and five miles wide, approximately was a blowout during light interven- than 0.4, would a single relief well be
10,000 ft below the surface of the sea tion. Scenarios 4 and 5 were blow- sufficient to kill the blowout. Early
and 6,600 ft below seabed level. outs to the seabed during drilling with development with big-bore wells will
Anticipated production capacity is 133/8-in. casing set at 6,890 ft TVD and have to be considered carefully in
approximately 38,000 to 53,000 B/D with 95/8-in. casing set right above the this case. JPT

1–2 April 2008


The Woodlands Waterway Marriott
Hotel & Convention Center
Houston, Texas, USA

Register Now!

w w w.s p e .o r g /c twi0 8 Society of Petroleum Engineers


www.spe.org www.icota.com

JPT • JANUARY 2008 75

You might also like