Casing/Tubing design manual october 2005 Chevron
Подождите немного. Документ загружается.
11.3.7 Use of VIT
Heat transfer from the production flow to well casings will be effectively reduced
when VIT is used. It then reduces the casing annulus pressure build up
accordingly, as the thermal expansion of any “trapped” annular fluids in the wells
is reduced. However, the use of VIT will greatly increase tubing cost.
11.4 WELLCAT
™
Modeling of Pressure
Build Up
WELLCAT
™
software can be used to model the pressure build up from
production for differing well configurations to asses the magnitude and effect of
TAPB. The following two cases are modeled to illustrate the range of likely
scenarios.
Case 1
Trapped annuli on a well (all annuli closed, except 22-in. annulus open to the
seabed) with no IT (example used is GC640#1 well). The modeling result shows
the following problems:
• Production will cause the annuli to develop very large trapped annular
pressures because of temperature elevation.
• Increase in annulus pressures will cause casing failures in this sequence:
o Burst of 22-in. casing
o Collapse of 17-7/8-in. casing
o Possible burst of 14-in. casing
Figure 11-2. Case 1 Before Production
Casing/Tubing Design Manual 11-7
October 2005
Figure 11-3. Case 1 After Production without VIT
Case 2
Trapped annuli on a well (all annuli closed, except 22-in. annulus open to the
seabed) with VIT (example used is GC640#1 well). The modeling result shows
the previous problems are no longer present:
• Production through the VIT string will greatly reduce the resultant trapped
annular pressures because of reduced temperature elevation,
• Increase in pressures will not cause casing failures.
\Figure 11-4. Case 2 Before Production
11-8 Casing/Tubing Design Manual
October 2005
Figure 11-5. Case 2 After Production with VIT
11.5 References
1. Moe, Bob and Erpelding, Peter: “Annular Pressure Buildup: What It Is and
What to Do About It”, Deepwater Technology (2000).
2. Leach, Colin P. and Adams, Adrian J.:“A New Method for the Relief of
Annular Heat-up Pressures”, paper SPE 25497 presented at the 1993
Production Operations Symposium, Oklahoma City, Oklahoma, March.
3. Adams Adrian J.: “How to Design for Annulus Fluid Heat-up”, paper SPE
22871, presented at the 1991 Annual SPE conference in Dallas, Texas,
October.
4. MacEachran, Angus and Adams, Adrian J.: “Impact on Casing Design of
Thermal Expansion of Fluids in Confined Annuli,” paper SPE/IADC 21911
presented at the 1991 Annual SPE/IADC Drilling Conference in Amsterdam,
Netherlands, March.
5. MacEachran, Angus and Adams, Adrian J.: paper SPE 29229 supplement to
above 21911 – same title and authors (1994).
6. “Viewpoint on Sustained Casing-head Pressure on Subsea Well Casing
Annuli,” FMC Energy Systems, Houston (October 2002). The FMC contact is
Brian Skeels.
7. “Practical Successful Prevention of Annular Pressure Buildup: The Marlin
Project” Faul, Ronnie and Vargo, Richard. (Discusses Nitrogen foamed
spacers) Contact Ronnie Faul at Halliburton.
8. Patillo P.D. and N. C. L: “Thermal and Mechanical Considerations for the
Design of Insulated Tubing,” paper SPE 79870.
Casing/Tubing Design Manual 11-9
October 2005
12
12Probabilistic Casing Design C
12.1 Introductio
oncept
n ....................................................................................................................12-1
12.2 Probabilistic Casing Design Approaches........................................................................12-3
2.1
2.2
12 o
12.1 Introduction
ied and discussed in recent years to
optimize casing design for safety and cost savings. Although the concept of
se
ct a
probabilistic casing design that calculates the probabilistic distributions of
We will the traditional (deterministic) casing design, from which
the casing design is conducted, by using the minimum expected strength
* Casing Load (12-1)
The de e
industry and is relatively easy to use, though:
•
r because of improved casing
The e maximum, while the actual casing load may be
lower because of the difficulty of casing load estimation.
12 Casing Strength Probability Design...........................................................................12-3.
12. Casing Load Probability Design ................................................................................12-6
.3 ad and Resistance Factor Design (LRFD)................................................................12-10 L
.4 Probabilistic Casing Design Applications .....................................................................12-1012
12.5 References...................................................................................................................12-11
Probabilistic casing design has been stud
probabilistic casing design is sound, probabilistic casing design is not easy to u
and has not become the main design method in the oil industry. This is because
the probabilistic casing design requires much more work for these reasons:
• A database must be built for casing load and casing strength.
• New casing design software or analysis must be used to condu
casing load and casing strength and designs the casing with certain
probability (risk).
first take a look at
(capacity) to exceed maximum possible loading, usually with a design factor
added “safety” as a contingency (see Figure 12-1):
Casing Strength (Capacity) >= Design factor
terministic (traditional) casing design has been used successfully in th
• The design factor is generally based on past design experience and doesn’t
really indicate the safety of the design.
The casing strength (capacity) is considered the minimum, while the actual
casing strength (capacity) may be highe
manufacturing technology.
casing load is considered th
Casing/Tubing Design Manual 12-1
October 2005
Load or Capacity Magnitude
Load
Capacity
"Safety" by
Design Factor
"Safety" by Design Factor = Capacilty/Load
Figure 12-1. Deterministic (Traditional) Casing Design
Probabilistic casing design, however, establishes the probability distributions of
casing load and casing strength (capacity), and then designs the casing with
certain probability (risk) of casing failure. The failure is calculated from the casing
load distribution is exceeding the casing strength (capacity) or from the casing
load probability distribution and the casing strength (capacity) probability
distribution is overlapped (see Figure 12-2).
Load or Capacity Magnitude
Probability
Density
Load
Capacity
Risk = Probability {Load > Capacity}
Risk
Figure 12-2. Probabilistic Casing Design
The casing load probability distribution is a combination of the probability
distributions of several loading components, such as pore pressure, kick volume,
and mud weight. The casing strength (capacity) probability distribution is a
combination of the probability distributions of the several casing
material/dimension components, such as yield strength, wall thickness, and pipe
ovality (see Figure 12-3).
12-2 Casing/Tubing Design Manual
October 2005
Load or Capacity Magnitude
Probability
Density
Load
Capacity
Risk
Yield strength
Wall thickness
Ovality
Mud weight Pore pressure
or kick volume
Figure 12-3. Determination of Casing Load and Casing Strength Probability
Distributions
Therefore, certain amount of sampling or testing plus a certain degree of
probability analysis has to be performed to obtain the probability distributions of
the casing material/dimension components (yield strength, wall thickness, pipe
ovality, etc.), and then calculate the combined probability distribution of casing
strength (capacity). Similarly, certain amount of previous drilling data has to be
analyzed to obtain the probability distributions of the loading components (pore
pressure, kick volume, mud weight, etc.), to calculate the combined probability
distribution of casing load.
12.2 Probabilistic Casing Design
Approaches
12.2.1 Casing Strength Probability Design
This probabilistic casing design approach is to improve the casing strength
(capacity) estimation by considering the probability distribution of the casing
strength (capacity) on improved understanding and accurate qualification of
casing materials and dimensional casing properties through testing and
probability analysis. The casing load is still to be considered the “maximum
casing load” as in detrimental casing design and the design equations is still to
use the same format as the in detrimental casing design:
Design factor * Casing Load = Casing Strength (12-2)
This probabilistic casing design was successfully used for a casing collapse
design for 7-in., 29-ppf, L-80, 13Cr casing in the South China Sea.
The API collapse rating on this casing is 7,030 psi. By testing forty-eight test
specimens, cut from twelve full-length joints selected at random from three
separate heats of initial manufacturing footage, the mean collapse failure
pressure (µ) from the test results was 9,114 psi with a standard deviation (σ) of
222 psi. With assuming a normal distribution to the casing collapse test data, the
following equation was used to qualify the improved casing collapse strength,
which limits the probability that the collapse strength is below 1%.
Casing/Tubing Design Manual 12-3
October 2005
P
c
=
−
µ
σ
2574.
where µ is the mean value and σ is the standard deviation of the overall test
collapse failure pressure data.
The collapse strength (capacity) of the 7-in., 29-ppf, L-80, 13Cr casing was then
raised from the API rating of 7,030 psi to the test-qualified 8,542 psi (see Figure
12-4). This 20+% margin between the API rating, 7,030 psi, and the test-qualified
strength, 8,542 psi, implies that the manufacturing processes and quality control
procedures for this product resulted in a consistent, high-quality product for which
higher performance properties were justified. This casing collapse capacity was
then used to design the casing string with the detrimental (traditional) safety
factor and casing load to realize substantial cost savings.
Collapse Pressure Test Results of 7", 29 ppf, L-80, 13Cr
Probabilit
y
Density
A
PI Ratin
g
7030 psi
σ
= 222
Collapse testing results
µ
= 9114
9114
7030
Qualified ratin
g
8542 psi
Figure 12-4. Qualification of Casing Collapse Strength by Probability Analysis
Another example is a casing burst design for a 4,000-foot-deep well with 5,000-
psi maximum anticipated surface pressure (MASP). The 5.5-in., 17-ppf, N-80
was initially recommended by API burst rating. However, when using the casing
strength probability design, a lower grade of 5-in., 17-ppf, J-55 casing was finally
selected. With assuming normal distributions, the mean and standard deviation
of wall thickness, yield strength, and tensile strength of the 5-in., 17-ppf, J-55
casing are calculated directly from mill data and shown in Table 12-1.
Based on the statistics of the casing properties, the distributions and standard
deviation of the casing burst strength according to the Barlow, Lame, and Plastic
Failure Pressure (PFP) criteria were calculated and presented in Table 21-1
12-4 Casing/Tubing Design Manual
October 2005
Table 12-1. With calculating the statistic burst strength by the (m - 2.574 s) value
for each of these criteria, the resulting design factors are all larger than 1.30 for
this well under the maximum anticipated surface pressure (MASP) of 5000 psi,
and indicates the adequacy of using the 17-ppf, J-55 tubular.
Casing/Tubing Design Manual 12-5
October 2005
Table 12-1. Pipe Properties and Burst Ratings for 5-1/2-Inch Casing
5-/2-in., 17-ppf, J-55 Nominal Mean Standard
Deviation
OD, in. 5.5 5.5 0
WT, in. 0.304 0.302 0.004
Yield, psi 55000 65000 2000
MASP, psi 5000
Nominal Mean Standard
Deviation
Barlow, psi 6080 7138 239
Lame, psi 6414 7527 254
PFP, psi 6621 7769 264
VME ID yield @ 50% yield
tension
6626 7786 260
API burst rating, psi 5320
At (µ-2.574σ)
Barlow Lame PFP VME ID yield
Burst rating, psi 6523 6873 7090 7117
Design factor 1.305 1.375 1.418 1.423
12.2.2 Casing Load Probability Design
This simplified probabilistic casing design approach is to work on the casing load
probability design centering on the management of geological uncertainty. The
following variables are considered to directly impact the design (an example well
data is shown):
• Depth of reservoir - The uncertainty is based on the error in seismic
interpretation. Figure 12-5 shows the depth uncertainty as a normal
distribution for the example probabilistically designed well, where the
maximum and minimum figures are calculated from the maximum possible
error in the conversion of the seismic data to depth. The means of
distribution is based on top of the reservoir. The maximum and minimum on
the normal distribution curve represent three standard deviations so the
distribution covers 99.6% of the possible depths.
• Reservoir pressure - The variation in the reservoir pressure is based on
data obtained from the offset wells. Figure 12-6 shows the uncertainty in the
example reservoir pressure as a triangular distribution. The expected figure
is the pressure encountered in the nearest offset well and the maximum and
minimum figures represent the highest and lowest pressures encountered in
similar highly over-pressured wells in the area.
• Reservoir fluid gradient (oil/gas gradient) – This is obtained from offset
wells based on modeling and burial history analysis. Figure 12-7 shows
uncertainty in the example reservoir fluid gradient as two combined normal
distributions with a 90% chance of oil and 10% chance of gas. The mean oil
and gas gradients are based on those encountered in the offset wells.
12-6 Casing/Tubing Design Manual
October 2005
Figure 12-5. Reservoir Depth Distribution with Expected Top Reservoir at 3,615 m
Figure 12-6. Distribution of Formation Pressure Gradient with Expected Mean at
20.05 kPa/m
Casing/Tubing Design Manual 12-7
October 2005