Casing/Tubing design manual october 2005 Chevron
Подождите немного. Документ загружается.
concentration, as long as the casing maximum tensile stress is less than 20% of
the casing material yield strength. This is roughly equivalent to a tension or burst
design factor of 5.00 to meet the sour cracking threshold tensile stress (20% of
the casing material yield strength) regarding the axial tensile stress and hoop
tensile stress. If this tensile stress threshold can be met in the steam injection
well, there should be no problem using P-110 and Q-125 grade casings,
regarding SSC.
Figure 10-7. Maximum Safe Stress Level for Various Grades of Casing and H
2
S
Concentrations at 24ºC (75ºF)
3
(Reference 4)
10.2.6.2 Thermal Wellhead
Thermal wellhead allows the production casing to expand and, therefore,
effectively avoid high thermal compressive load and casing buckling at the top
section of the production casing (Figure 10-8). This is where cement is likely
missed because of the cement fall-back at the end of the cementing. Therefore, it
can help reduce casing failure at the production casing top section because of
hot-yield, buckling, and collapse.
10-12 Casing/Tubing Design Manual
October 2005
Casing/Tubing Design Manual 10-13
October 2005
This approach can help to avoid the casing thermal compressive load later in
steam injection period and prevent casing hot-yield.
As shown in Figure 10-9, when 300,000 lb pre-tension is applied to the 7-in., 29#,
L-80 production casing in the example well, the casing thermal compressive load
(under steam injection condition 288ºC (550ºF) injection temperature) is reduced
accordingly. The 7-in., 29#, L-80 production casing will not be hot-yielded (casing
load line stays inside the casing von Mises yield ellipse) under steam injection
conditions.
• Casing lands with pulled tension at the casing top when the quick-set cement
at the bottom section of the casing can hold the amount of pulled tension.
• Quick-set cement slurry is pumped to the bottom section of the production
casing
The following actions take place during the casing pre-tension completion:
10.2.6.3 Casing Pre-Tension Completion
Figure 10-8. Thermal Wellhead to Allow Production Casing Expansion (Courtesy of
Cameron)
Figure 10-9. Pre-Tension Production Casing Helps to Avoid Casing Hot-Yield during Steam Injection
Casing/Tubing Design Manual 10-14
October 2005
10.3 Other Issues
10.3.1 Casing Connection
High-compression efficiency casing connection is needed in steam injection wells
because of the severe well conditions of high temperature, extreme axial
compressive load, and possible H
2
S environment. A premium connection with
torque shoulder and metal-seal will perform much better than the common API
BTC and LTC connections. Although API BTC and LTC connections are low
cost, they may present a high risk of connection failure.
10.3.2 Connection Lubricant
High temperature connection lubricant has to be used to avoid connection leaks
under steam injection conditions.
10.4 References
1. Lepper, G.B.: “Production Casing Performance in a Thermal Field,” JCPT,
(1998).
2. Maruyama, K. et al.: “An Experimental Study of Casing Performance under
Thermal Cycling Conditions,” SPE Drilling, (1990).
3. Bourgoyne Jr. A. T. et al.: “Applied Drilling Engineering,” Vol. 2, SPE
Textbook Series, Richardson, Texas, (1986).
4. Wu, J.: “Steam Injection Casing Design,” paper SPE 93833, presented at the
2005 SPE Western Regional Meeting, Irvine, California, 30 March–1 April.
Casing/Tubing Design Manual 10-15
October 2005
11
11Casing Design for Deep Water Wells
11.1 Introduction ....................................................................................................................11-1
11.2 Sustained Casing Pressure............................................................................................11-1
11.3 Alternatives for Mitigation of TAPB.................................................................................11-2
11.3.1 Leave Cement Short of the Previous Shoe ...............................................................11-2
11.3.2 Use Nitrogen (N2) Ahead of the Cement Slurry ........................................................11-3
11.3.3 Install a Crushable Foam on the Outside of the Casing String..................................11-4
11.3.4 Use of Burst/Collapse Discs in Conductor and Surface Casing Strings ....................11-5
11.3.5 Alternatives for Surface Casing.................................................................................11-6
11.3.6 Use of Load-Resistant Casing...................................................................................11-6
11.3.7 Use of VIT .................................................................................................................11-7
11.4 WELLCAT
™
Modeling of Pressure Build Up...................................................................11-7
11.5 References.....................................................................................................................11-9
11.1 Introduction
Casing design for deepwater wells needs to deal with an additional problem -
excessive annulus pressure build up because of thermal expansion of annular
fluids in the wells.
In high productivity or high temperature oil wells, the heat transfer from the
production tubing increases the temperature in the outer well annuli. If the fluid in
those spaces is unable to expand (i.e., is “trapped”), then significant pressure
build up occurs (in the range of 7,000 to 9,000 psi), which if not included in the
design, can cause casing, tubing, and well failures. This is referred to as thermal
annulus pressure buildup (TAPB). On traditional platform wells, this expansion
has been routinely bled off during the initial days and weeks of a well production
life by bleeding the appropriate annuli side outlet valves. In subsea wells where
there is no method to either monitor or bleed fluid from outer annuli (other than
the tubing (A) annulus), there is the potential for very high pressures to build up,
which may not be contained by commonly available casing strings.
11.2 Sustained Casing Pressure
There are several sources of sustained casing pressure (SCP), as shown in
Figure 11-1. TAPB is a separate category, and has been known about in higher
temperature wells for many years. Several industry publications exist from the
early 1990s describing the issue and discussing how to handle it (see the
References section of this chapter). There have been a few high-profile well
failures as a result of this problem. An example includes the BP “Marlin” field in
2000 where the tubing in one well collapsed because of pressure build up in the
“A” annulus just hours after starting production. This caused extensive well
remedial work and the retrofitting of vacuum insulated tubing (VIT) at
considerable lost production and cost.
Casing/Tubing Design Manual 11-1
October 2005
02/17/2003
02/17/2003
MMSMMS
From “A Review
of Sustained
Casing Pressure
(SCP) Occurring
on the OCS” by
Bourgoyne et al.
(March 2000)
Mud cake
Casing
leak
Well-head leak
Tubing leak
Tensile crack in cement
caused by temperature
cycles
Underground blowout
Low pressure sand
High pressure sand
Channel caused by flow
after cementing
Micro-annulus caused
by casing contraction
Figure 11-1. Sources of SCP
11.3 Alternatives for Mitigation of TAPB
11.3.1 Leave Cement Short of the Previous
Shoe
If the top of cement (TOC) is left below the previous casing shoe when a string of
casing is cemented, then a natural “relief valve” is provided. Because, when the
annular fluid above the TOC is heated, its pressure will only build up to a level
that causes break down of the formation at the shoe. Then, the fluid leak offs to
the formation and pressure stabilizes at or near the fracture propagation
pressure. In most deepwater wells, the formation leak-off pressure is only slightly
above the hydrostatic in the annulus (because of the narrow pore
pressure/fracture gradient window). Therefore, annulus pressures typically do not
build up to more than 2,000 psi. For example, in an annulus with a shoe at
20,000 feet with a 15-ppg leak-off and 13-ppg mud, the annulus pressure should
be limited to 2,000 psi.
However, there are some issues to be aware of when using this method of
annulus pressure build-up control, particularly in directional wells. Because of the
following issues, this method cannot be guaranteed.
• There is a possibility of cement channeling up into the previous shoe and
creating a seal. For this reason, good centralization throughout the cemented
interval is important to improve cement displacement and minimize
channeling.
• The annulus could become plugged with solids settling out of the mud and
spacer above the cement. Typically, spacers are not capable of suspending
11-2 Casing/Tubing Design Manual
October 2005
their solids for longer than a few days and barite will also eventually sag in
the mud. There is, therefore, a chance that these solids could seal the
annulus above the previous shoe and create a pressure seal causing
excessive annular pressure build up when production begins later. As an
example, a well in the Petronius field had a TAPB problem even though the
well had a TOC below the shoe. Barite settling became trapped the annulus
in question.
For these reasons, even when cement is designed to be left below the previous
shoe, it is recommended that an additional secondary mitigation method be used.
11.3.2 Use Nitrogen (N2) Ahead of the Cement
Slurry
The principle is to place a small amount of compressible gas in the annulus to
allow the annular fluid to expand when it is heated, minimizing any additional
pressure build up. A relatively small amount of gas is required to accomplish this
and nitrogen (N2) is most suitable because it is inert and widely available and its
use is proven.
The easiest way to place the N2 is to add it to the cement spacer and pump a
foamed spacer. In this way it is mixed during the cementing process and is finally
positioned just ahead of the lead slurry. In time it will break out of the spacer fluid
and migrate up the annulus coming to rest at the top of the annulus. Herein lies
the main drawback with this method, because as it migrates and is unable to
expand, it:
• Brings the hydrostatic pressure from the depth where it was originally
positioned to surface.
• Exerts that pressure as an applied pressure on the annulus, and decreasing
the effectiveness of the method to reducing annulus pressure build up.
However, for shallow strings of casing, this is a preferred method. For deep
strings, this problem can also be overcome by pumping the foamed N2 earlier in
the cementing sequence at some predetermined volume ahead of the traditional
spacer and cement to position it higher up the annulus on completion of the
cement job. This minimizes the migration distance and trapped pressure from the
migrated N2.
This method is being increasingly used as a mitigation method. As a guideline,
5% N2 by volume at surface of the sealed annular volume is required to absorb
typical annular fluid expansion and a design factor of 2 is recommended, so 10%
of N2 is usually pumped. Foamed spacers can be designed to be stable for about
three to five days, giving ample time for casing operations with contingency for
problems setting wellhead seal assemblies, etc., before the N2 breaks out.
This method has been used extensively by BP on 13-3/8 in. and 9-5/8 in.
cementations on Marlin and Dorado fields and also in other wells. This is
discussed in detail in a Halliburton white paper entitled “Practical Successful
Prevention of Annular Pressure Buildup: The Marlin Project.” Contact Ronnie
Faul at Halliburton for more information. A recent deep sub-salt well in the Gulf of
Casing/Tubing Design Manual 11-3
October 2005
Mexico (Conoco’s Spa Prospect) also used this method for the 13-5/8-in x 22-in.
annulus (see SPE paper 79810).
This method does have some notable risks:
• If losses are experienced during the cementation, the N2 may be lost to the
formation or may be left very deep in the annulus.
• Excessive channeling could result in the N2 coming to rest in the marine riser
above the wellhead eliminating it as a TAPB prevention and potentially
causing problems in having to circulate it out of the well.
• The depth of placement must be optimized to limit the pressure increase as it
migrates to the wellhead.
The method is recommended as a primary mitigation for use in the production
tieback annulus on all Tahiti wells. However, a secondary method should also be
designed into the well because of the potential problems described above.
Additionally, this method is recommended as the secondary method of mitigation
for all intermediate casing strings.
11.3.3 Install a Crushable Foam on the Outside
of the Casing String
Syntactic foam with a specific collapse pressure is available for installation onto
the outside of the casing in quadrant sections just below the subsea wellhead.
The required collapse pressure must be specified so that on installation it does
not collapse. However, when annulus pressure subsequently builds up, it
collapses at some value higher than hydrostatic but still safely below the rating of
the surrounding casings (say, 1,500-psi higher than hydrostatic).
CRP Marine in the U.K., (with an office in Houston, Texas) is a supplier of this
material, and has previously provided details and a quote to provide foam and
installation services for this kind of application. The cost would be in the order of
U.S. $50 to $100,000 per casing string depending on the volume of foam
required.
The issues with this method and product are mainly related to a reliable
installation and track record. The foam is glued onto the outside of the casing in
the pipe yard and banded for additional security. There is a concern of damage
during handling and transit and during running from the rig floor through the riser/
BOP/wellhead. The possibility of some foam breaking off and being in the
wellhead when attempting to land a subsea hanger and set a pack-off seems a
real threat and risk. Although the product was tested in a well by BP, it has not
been pursued as one of the preferred methods for solving this problem and CRP
Marine does not seem to be active in promoting its use for this purpose.
This method is not recommended for use in any of the casing strings in Tahiti
wells at this stage. There is an advanced method of placement of syntactic foam
that is being developed for BP wells. In this application, the foam is molded as
part of the casing below the connection upset. BP has had success with this type
of strategy. We are currently looking at this as an alternative for future Tahiti
wells.
11-4 Casing/Tubing Design Manual
October 2005
11.3.4 Use of Burst/Collapse Discs in
Conductor and Surface Casing Strings
Although surface casings are typically designed to be cemented to surface, it is
considered almost inevitable that a small amount of the annulus space at the top
will be left with fluid inside. This is because the heavy cement is likely to leak off
a little at the casing shoe and fall back in these shallow, porous weak formations
leaving fluid in the top of the annulus. In addition, gravity segregation of the
solids will always leave liquid on top. For these reasons, if the 22-in. x 36-in. (or
38-in.) casing is sealed, TAPB will most likely occur.
Burst disc designs are available and have been used on conductor and surface
casing strings. These discs are uni-directional and will fail at specified pressure
from the direction of designed load, but will withstand typically five times that
pressure from the opposite direction. Other specifics about the discs include:
• They are contained within a plug threaded into a tapped hole in the wall of
the casing.
• They are designed to rupture from one direction at a predetermined
differential pressure, which is specified to be less than the casing burst (or
collapse).
• They are a precision disc, manufactured to any specified rating and are
accurate to within 5% of that rating.
• They are typically specified to 90% of the rating of the casing in question,
meaning that the failure pressure should be between 85% and 95% of the
casing rating. In some cases it may be necessary to use discs to protect from
burst or collapse loads or both. This can be accommodated by fitting
separate discs rated for each application in the correct orientation.
The main drawback of this method is the possibility of early inadvertent rupture. If
this occurs during the drilling of the well in a pressure containing casing (i.e.,
after BOP installation) before the next string is set, then well integrity could be
lost. Indeed, this is reported to have happened in one confirmed case at BP. It
was caused by human error (an operator put his screwdriver through the burst
disk).
Therefore, at this time, use of burst discs are only recommended in the outer
casing strings (20 in. and above), and where a significant differential pressure
could not be applied during the spudding of the well (because of operating
riserless in open water). However, after the 22 in. is installed, a trapped annulus
is possible and the burst disc will protect from 36-in. burst or 22-in. collapse. The
burst disc rating would be determined from the lowest of the conductor burst and
22-in. collapse rating.
A supplier who has provided discs for use on BP wells is Alpha Process Sales
Inc., based in Sugar Land, Texas, (281-240-4505). This company supplies discs
made by the long-established Fike Corporation who traditionally manufactures
rupture discs for the process/pipe/vessel protection market. Another supplier is
Hunting Tubular Systems. Hunting has excellent QA/QC systems to ensure
human error is minimized and they are BP’s main suppliers of burst discs.
Casing/Tubing Design Manual 11-5
October 2005
11.3.5 Alternatives for Surface Casing
Instead of burst discs, there are a couple of alternative ways to mitigate the
above problem in surface casing strings. For the 22-in. x 36-in casing annulus, a
sealed annulus is usually created after cementing the 22 in. by setting the
annulus shut-off sleeve and by using an ROV to close the ball valves on the
return ports on the LP housing. This is usually done to prevent the possibility of
shallow flows exiting this annulus. It also creates a potential TAPB problem.
To solve this problem, on start up of production from such a well, the ROV-
operated valve could be opened to bleed pressure for a few days until the well
has thermally stabilized. Then, the valve can be closed again to protect against
shallow flows. In the event that the valve has corroded or seized in the time
between drilling the well and bringing it on production, the ROV could saw off the
valve or the annulus could be left alone.
Leaving this annulus sealed could also be another option (although not
recommended). If excessive pressure builds up, then the 22-in. casing could
collapse (depending on the pressure between the 17-7/8 in. and 22 in.), or the 36
in. could burst. Although undesirable, rupture of the 36 in. is not catastrophic,
because at this time in the well’s life it carries only bending loads and this would
not be significantly affected by a short split in the casing. Collapse of the 22 in.,
however, could have a chain reaction affect on other inner casings and be
catastrophic.
11.3.6 Use of Load-Resistant Casing
In most cases, casing can be designed to withstand the calculated loads from
TAPB. However, although this may be possible and appropriate in some well
designs where the anticipated loads are estimated to be reasonable, it is
impractical in many wells and it is also fraught with potential problems. It may
drive the design to excessive wall thicknesses, high-yield strengths, and low
safety factors.
This presents problems in terms of casing cost, smaller annular clearances (high
ECDs and swab/surge), heavy casing running weights, and more risk (and less
tolerance to unknown factors). This last risk factor is probably the most serious
because a number of assumptions are made in casing design calculations and
multi-string analyses, which may not turn out to be accurate in the field.
Additionally, any number of minor issues could be unforeseen or overlooked,
which could turn out to have enough of an effect on a low safety factor to allow
casing failure. Where safety factors are low, then the risk of casing failure
because of these unknowns is a real one. It is considered a much more prudent
practice to try to lower the loads on casing rather than construct a well to
withstand high loads. For this reason, the approach taken here is to find ways to
limit TAPB and then design the casing to meet whatever is the driving load case,
rather than accept high TAPB as the driving load case for the well design.
11-6 Casing/Tubing Design Manual
October 2005