Document No.
DWGOM-GP 02-1019
Applicability
DWGOM
Date
Draft
New Criterion for Jetting/ Conductor Design
DWGOM-GP 02-1019
DWGOM SITE TECHNICAL PRACTICES
BP 2002 UTG DRILLING TECHNOLOGY
Advanced Well Design Assurance Project
New Criterion for Jetting/Conductor Design
BP 2002 UTG DRILLING TECHNOLOGY
Advanced Well Design Assurance Project
BP America Inc. 501 Westlake Park Boulevard Houston, Texas 77079
2 0 0 2
B P
U T G
D R I L L I N G
T E C H N O L O G Y
P R O J E C T
CONTENTS Introduction................................................................................................................. 1 B.1
Conductor Capacity ...................................................................................... 2 B.1.1 Immediate Capacity....................................................................................2 B.1.2 Effects of Soil Set-up On Capacity...........................................................3 B.1.3 Total Load Capacity of the Jetted Conductor .........................................4
B.2
Conductor Design ......................................................................................... 6 B.2.1 Selection of Soil Profiles for Gulf of Mexico Deepwater Sites .............6 B.2.2 Landing Load...............................................................................................7 B.2.2.1 Volume Displacement Method..................................................... 7 B.2.2.2 Pressure-Area Method.................................................................. 8 B.2.3 Length of Conductor.................................................................................10 B.2.4 Weight On Bit Profile................................................................................12 B.2.5 Reciprocations ..........................................................................................15
B.3
Recommendations ...................................................................................... 18 B.3.1 Design Procedure .....................................................................................18 B.3.2 Documentation ..........................................................................................18 B.3.3 Operations .................................................................................................19 B.3.4 Using JETCAP ..........................................................................................20
B.4
Well Failures ................................................................................................ 21
B.5
Glossary of Terms ....................................................................................... 22
B.6
References ................................................................................................... 23
APPENDIX A.............................................................................................................. 24 Derivation of the Required Conductor Length, L................................................24 APPENDIX B.............................................................................................................. 27 Example Calculation: Landing Load ...................................................................27 APPENDIX C.............................................................................................................. 29 Example Calculation: Required Conductor Length...........................................29
Draft D, 12/13/02
DWGOM-GP 02-1019
B-i
2 0 0 2
B-ii
B P
U T G
D R I L L I N G
T E C H N O L O G Y
P R O J E C T
DWGOM-GP 02-1019
Draft D, 12/13/02
2 0 0 2
B P
U T G
D R I L L I N G
T E C H N O L O G Y
P R O J E C T
INTRODUCTION The importance of a sound structural foundation cannot be overemphasized. In the past decade, there has been a significant amount of research regarding the design and installation of jetted conductors. Yet, despite the lessons learned, seven conductors on BP wells subsided in 2002. In each of these cases, at least one of the recommended design/installation techniques presented in this section was ignored. The earliest jetted conductors were designed by sampling and analyzing Gulf of Mexico (GOM) soil properties. Using these soil analyses, lower and upper bound immediate capacities for conductor setting depths were calculated1. The designs were then configured for the length of the conductor casing, application of donut weights, and bottom hole assemblies. Reciprocation was avoided by having enough weight available during jetting. This was achieved by the use of donut weights, which were cylindrical weights of steel that were stacked up on the drill pipe. The key advantage of this method was that the weight of the donuts used was greater than the weight to be landed on the conductor. If the conductor was able to carry such weight immediately after installation, then the conductor was essentially proof-loaded against the weight to be landed. This provided certainty that the wellhead would not move when landing of subsequent strings took place. The main limitation of this method was that it was not possible to drill ahead after jetting was completed. The BHA had to be retrieved, the donut weights removed, a new BHA assembled, and drill pipe run back in the hole again before drilling could resume. As water depths continued to increase, this disadvantage became such a significant cost factor that this method was eventually abandoned. The main issue at hand was that none of these earlier practices calculated the increase in unit friction along the conductor with time, nor did they sufficiently account for soil properties. Well failures were not adequately studied to determine the causal factors and relationships (e.g., conductor penetration lengths or reciprocation stroke lengths). Considering the high cost of drilling deepwater wells, these and other factors needed to be addressed. Through various studies and collaborations, BP recommends this standardized approach for the design, setup, and installation of deepwater conductors. It is now no longer common practice for BP to collect site-specific soil data to determine conductor setting depths for exploration wells in the GOM. Therefore, the proposed method for setting depth design of a conductor assumes reasonable conditions. The BP model has been strengthened by the lessons learned from failure studies, and the final result reduces both setup time and material cost.
Draft D, 12/13/02
DWGOM-GP 02-1019
B-1
2 0 0 2
B.1
B P
U T G
D R I L L I N G
T E C H N O L O G Y
P R O J E C T
CONDUCTOR CAPACITY Setup calculations have recently been facilitated by the implementation of the Jetting Capacity (JETCAP) program and by the recent studies performed on various GOM deepwater wells. BP has determined that donut weights are too time consuming and, therefore, too costly. In order to determine conductor capacity, the following recommendations and equations are presented.
B.1.1 IMMEDIATE CAPACITY The basic make-up of the casing drill assembly includes a bit, motor, jet, crossover and spacer subs, and drill collars. These are attached to a running tool, which is latched to the conductor assembly. The inside string is spaced according to bit position relative to the conductor shoe such that the jetting nozzles are located 12” to 18” above the conductor tip. The conductor also adds weight to the bottom hole assembly (BHA), which further aids in the jetting process. The immediate capacity of the conductor is essentially determined by the weight on bit (WOB). Physically, the last soil resistance measured during installation should be equal to the immediate capacity. Maximizing the recorded WOB during jetting is of prime importance; therefore, the last WOB should be equal to at least 80% of the available WOB during jetting. The available WOB is calculated by adding the buoyant weights of the conductor, the wellhead housing, the drill collars, and the Cam-Actuated Drill Ahead (CADA) tool. This is expressed as: Q0 = WOBlast = (R )(Wcond + WWH + WDC + WCADA)
(B.1)
where: Q0 = Conductor capacity immediately after jetting (t = 0.01 days) WOBlast = Last weight on bit recorded during installation R = WOB utilization ratio, should be between 0.8 and 1.0 Wcond = Weight of the conductor in water WWH = Weight of the wellhead housing in water WDC = Weight of the drill collars in water WCADA = Weight of the CADA tool in water
B-2
DWGOM-GP 02-1019
Draft D, 12/13/02
2 0 0 2
B P
U T G
D R I L L I N G
T E C H N O L O G Y
P R O J E C T
B.1.2 EFFECTS OF SOIL SET-UP ON CAPACITY BP’s past practice was to increase the immediate capacity by 20% per log cycle of time so that after 10 days the capacity was estimated to have increased by a factor of 1.6. In order to improve on this technique and make it applicable for a wider range of conditions (i.e., soil profile), conductor diameter, and length; the factor Dat was developed:
∆αt =
Qt - Q0 (π)(D )( L )( Suave )
(B.2)
where: Dat = Setup factor; measures the change in average friction factor along the conductor, at time = t
Q0 = Initial capacity, (at time t=0.01 days) Qt = Conductor capacity at time = t days < 10 days D = Conductor diameter L = Conductor length below mudline Suave = Average undrained shear strength over the length of the conductor Suave is the average undrained shear strength of the clay over the embedded length L, as determined by a best fit through available data. In order to develop the Dat relationship with time, the existing database of load tests on jetted conductors was used. A sample of the data used is given in Table B-1 and plotted on Figure B-1.
Site
Conductor Length (ft)
Conductor Diameter (in.)
Immediate Capacity (kips)
Setup Time
Measured Capacity (kip)
Calculated Dat
A
135
30
135
4 hr
>170
0.07
B
65
30
91
1 day
125
0.11
C
175
36
130
5 days
>300
0.15
Table B-1–Summary of Load Test Results on Partially Setup Jetted Conductors
Draft D, 12/11/02
DWGOM-GP 02-1019
B-3
2 0 0 2
B P
U T G
D R I L L I N G
T E C H N O L O G Y
P R O J E C T
0.18 0.16 0.14
Calculated Da
0.12 0.1
Data Best Fit
0.08 0.06 0.04 0.02 0 0.01
0.1
1
10
Time (days)
Figure B-1–Measured Setup Factor as a Function of Time
Plotting the line through the data in Figure B-1 results in the following equation: Dat = (0.055)[2 + log(t)]
(B.3)
Immediately after jetting the conductor, the soil in the immediate vicinity around the conductor begins to set up and regains strength, resulting in an increased capacity of the conductor with time. The Da is a measure of this healing process and can be used to predict the capacity of the conductor as a function of time after jetting.
B.1.3 TOTAL LOAD CAPACITY OF THE JETTED CONDUCTOR Substituting Eqs. B.1 and B.3 into Eq. B.2, and rearranging, give the following expression for the total conductor capacity: Qt = WOBlast + (0.055)[2 + log(t)](p)(D)(L)(Suave)
(B.4)
where: t = time, in days < 10 days
B-4
DWGOM-GP 02-1019
Draft D, 12/13/02
2 0 0 2
B P
U T G
D R I L L I N G
T E C H N O L O G Y
P R O J E C T
Qt = Conductor capacity at time = t days WOBlast = Weight on bit measured at final penetration D = Conductor diameter L = Conductor length below mudline Suave = Average undrained shear strength over the length of the conductor Eq. B.4 represents a lower bound estimate, because the data points used to develop it were based on proof load tests, and not on load tests to failure. This method is likely to be conservative, particularly for short setup times (i.e., less than one day). However, it is not recommended to increase the setup rate until new data becomes available. In addition, Eqs. B.3 and B.4 should not be used for setup times greater than 10 days.
Draft D, 12/11/02
DWGOM-GP 02-1019
B-5
2 0 0 2
B.2
B P
U T G
D R I L L I N G
T E C H N O L O G Y
P R O J E C T
CONDUCTOR DESIGN
B.2.1 SELECTION OF SOIL PROFILES FOR GULF OF MEXICO DEEPWATER SITES It is not common practice for BP to collect site-specific soil data for exploration wells in the GOM. Therefore, the proposed method for designing the conductor must assume reasonable site conditions. BP has access to 15 high quality soil profiles that were obtained by dedicated geotechnical vessels on high quality 3-in. soil samples or by 4-in. jumbo piston cores. The data fits in a somewhat narrow range. The upper bound corresponds approximately to a rate of increase of 10 psf per ft of penetration, whereas the lower bound corresponds approximately to a rate of increase of 6.3 psf per ft rate of increase. To calculate the conductor capacity as per Eq. B.4, the average shear strength over the length of the conductor is required. Figure B-2 represents three profiles of average shear strength, from the mudline to a given depth, as a function of depth. The results are summarized in Table-B-2. In the absence of site-specific boring, and if the available seismic data does not show active or relic large scale geologic processes capable of influencing the shear strength profile (i.e., faulting, landslides, channeling), the use of the average shear strength profile is recommended for design.
Su AVE = Su 0 + Su1 z
Suo (ksf)
Su1 (ksf/ft)
Lower bound profile
-0.0268
0.0034
Average profile (recommended for design)
0.0191
0.0043
Upper bound profile
0.0635
0.0052
Table B-2–Determination of Average Shear Strength along the Conductor Length 0
0.2
0.4
0.6
0.8
1
1.2
1.4
1.6
1.8
0
Lower
50
Ave
Depth z (ft)
100
Upper
150 200 250 300 350
Figure B-2–Average Shear Strength from Mudline to Depth z (ksf)
B-6
DWGOM-GP 02-1019
Draft D, 12/13/02
2 0 0 2
B P
U T G
D R I L L I N G
T E C H N O L O G Y
P R O J E C T
B.2.2 LANDING LOAD The first step in determining the conductor length is to determine the load to be landed on it. The mud weights inside the hole, inside the surface casing, and inside the cementing string, as well as the weight indication on the drill pipe running string, all influence the actual landing load.
B.2.2.1 VOLUME DISPLACEMENT METHOD The landing load can be calculated by finding the force imbalance between the structural and fluid column loads acting down resisted by the weight of the fluid displaced by the outside of the surface casing. Figure B-3 along with Eqs. B-5 and B-6 illustrates this technique.
ΣFy = 0 = Wlanded – Wsurf – Wcmt – Wcmt-fld – Wcmt-surf-fld + Wsurf-fld
(B.5)
Wlanded = Wsurf + Wcmt + Wcmt-fld + Wcmt-surf-fld - Wsurf-fld
(B.6)
where: Wlanded = landing load Wsurf = air weight of surface casing Wcmt = air weight of cementing string (the drill pipe below the mudline) Wcmt-fld = weight of fluid inside cementing string Wcmt-surf-fld = weight of fluid occupying the volume between the outside of the cementing string and the inside of the surface casing. Wsurf-fld = weight of fluid displaced by the surface casing This formulation assumes that the weight of the cementing string is part of the landing load.
Draft D, 12/11/02
DWGOM-GP 02-1019
B-7
2 0 0 2
B P
U T G
D R I L L I N G
T E C H N O L O G Y
P R O J E C T
Wlanded Fluid inside surface casing
Wsurf
Wcmt Fluid outside surface casing
Fluid inside cementing string
Figure B-3–Volume Displacement Method to Determine the Landing Load
B.2.2.2 PRESSURE-AREA METHOD For maximum analysis flexibility, the pressure-area/free body diagram approach is recommended for calculating the landing load. Figure B-4 along with Eqs. B-7 and B-8 illustrates this technique.
ΣFy = 0 = Fhook – Wdp – PtAt + PbAb + Wlanded + PxoAxo – Wsurf – PziAzi – PziAdp + PzoAzo
(B.7)
Wlanded = -Fhook + Wdp + PtAt + PbAb - PxoAxo + Wsurf + PziAzi + PziAdp - PzoAzo
(B.8)
It is important to note that areas At and Ab are not equal. At includes the cross-sectional area at the top of the surface casing, whereas Ab only considers the area exposed between the inside of the surface casing and the outside of the drill pipe.
B-8
DWGOM-GP 02-1019
Draft D, 12/13/02
2 0 0 2
B P
U T G
D R I L L I N G
T E C H N O L O G Y
P R O J E C T
Fhook
+ PtAt
Wlanded
PbAb
Wsurf
Wdp
PxoAxo
PziAzi
PzoAzo
PziAdp
Figure B-4–Pressure-Area Technique to Determine the Landing Load with Drill Pipe
The influence of the drill pipe can be removed from the analysis in order to obtain a simplified solution for the landing load. An example landing load calculation is included in Appendix B.
ΣFy = 0 = – PtAt + PbAb + Wlanded + PxoAxo – Wsurf – PziAzi + PzoAzo Wlanded = PtAt - PbAb - PxoAxo + Wsurf + PziAzi - PzoAzo
Draft D, 12/11/02
DWGOM-GP 02-1019
(B.9) (B.10)
B-9
2 0 0 2
B P
U T G
D R I L L I N G
T E C H N O L O G Y
P R O J E C T
+ PtAt
Wlanded
PbAb Wsurf
PxoAxo
PziAzi
PzoAzo
Figure B-5–Pressure-Area Technique to Determine the Landing Load without Drill Pipe
When computing the landing weight, consider designing for the case during the cement job when the lighter weight spacer fluid is in the open hole and the cement is in the inner cement string, if appropriate.
B.2.3 LENGTH OF CONDUCTOR Once the surface casing landing weight has been calculated, the length of the conductor can be determined. It is very important to use a consistent set of units when using Eq. B-11. The derivation of Eq. 11 is made in Appendix A, and an example calculation is made in Appendix C.
L=
- B + B 2 - 4 AC + (4 A)( FS2 )(Wlanded ) 2A
(B.11)
where: L = Required conductor length below mudline A = (0.055)[2 + log(t)](p)(D)(Su1)
B-10
DWGOM-GP 02-1019
Draft D, 12/13/02
2 0 0 2
B P
U T G
D R I L L I N G
T E C H N O L O G Y
P R O J E C T
B = (R)(wDC + w2) + (0.055)[2 + log(t)](p)(D)(Su0) – (FS1)(w2) C = (R)[WWH + WCADA + (L1)(w1 – w2)] – (FS1)(WWH) + (FS1)(L1)(w2 – w1) and: D = Conductor diameter L = Conductor length below mudline T = Time, in days Wlanded = Weight of casing string to be landed
FS1 = Safety factor on weight of conductor and wellhead FS2 = Safety factor on weight to be landed Wcond = Weight of conductor in water WWH = Weight of well head housing in water WDC = Weight of the drill collars in water WCADA = Weight of the CADA tool in water wDC = Weight per unit length of drill collars in water w1 = Weight per unit length of upper conductor section in water L1 = Length of upper conductor section (usually equal to 80 ft) w2 = Weight per unit length of lower conductor section in water R = WOB utilization ratio, should be between 0.8 and 1.0 Su0 = equals 0.0191 ksf for an average GOM soil profile Su1 = equals 0.0043 ksf/ft for an average GOM soil profile
Draft D, 12/11/02
DWGOM-GP 02-1019
B-11
2 0 0 2
B P
U T G
D R I L L I N G
T E C H N O L O G Y
P R O J E C T
The ultimate choice of safety factor remains with the user, but it is recommended that the setting depth be determined using a minimum safety factor of 1.3 against the weight to be landed (i.e., FS1=1.0 and FS2=1.3). The available GOM data suggests that Eq. B.11 may be valid for conductors installed both with and without reciprocation, provided the jet nozzles are 12” to 18” above the conductor tip and the reciprocation strokes are not excessive. The WOB utilization ratio, R, should be kept less than 1.0 to avoid compressive stresses in the BHA and running string, as well as to prevent buckling. The immediate capacity is arbitrarily defined at a time equal to 0.01 day (on the log cycle graph). This corresponds to roughly 14 minutes, and is consistent with past practices. This method is inherently conservative since the setup curves are lower bounds based on proof load tests and not on load tests to failure. In addition, any drag due to hole deviation and/or doglegs reduces the weight to be landed. However, the recommended average shear strength profile may or may not be conservative depending on the site location. Also, keep in mind that this method does not account for any hydraulic fracture requirements, which may dictate a deeper setting depth.
B.2.4 WEIGHT ON BIT PROFILE The recommended WOB profile vs. penetration depth should be generated such that reciprocation is started when the WOB reaches about half of the maximum usable WOB during jetting. The maximum usable WOB can be calculated using Eq. B.12. WOBusable = R[w1L1 + w2(L – L1) + wDCL + WWH + WCADA]
(B.12)
The WOB profile from the mudline to the point where reciprocation starts can be calculated using Eq. B.13. The shear stress term is reduced by one-half to compensate for the reduced friction between the conductor and the disturbed soil moving past the conductor. A reduction factor of 0.5 is generally a good representation of this phenomenon.
WOBno recip = πD
( 2z ) (Su
o
+ Su1 z )
(B.13)
Once the depth is reached (zrecip) where the WOB calculated from Eq. B.13 is equal to one-half the useable WOB, the WOB profile should increase linearly so as to reach the maximum allowable WOB at target depth. Eq. B.14 can be used to predict the depth where reciprocation should begin. Be very careful to use a consistent set of units when using Eqs. B.13 and B.14. An example is illustrated in Figure B-6.
zrecip =
B-12
-0.5Suo +
usable ) (0.5Suo )2 + ( Su1WOB πD
Su1
DWGOM-GP 02-1019
(B.14)
Draft D, 12/13/02
2 0 0 2
B P
U T G
D R I L L I N G
T E C H N O L O G Y
P R O J E C T
where: WOBno recip = Recommended WOB up to a depth of zrecip WOBusable = Maximum usable WOB during jetting z = Depth below mudline zrecip = Depth below mudline where reciprocation begins Recommended Weight on Bit (kips) 0
0
20
40
60
80
0.5*Max available WOB
100
120
140
Max available WOB
50
Depth (ft)
100
150
200 Note: Dashed line indicates no reciprocation Solid line indicates reciprocation
250 Figure B-6–Recommended WOB profile During Jetting Operation
Draft D, 12/11/02
DWGOM-GP 02-1019
B-13
2 0 0 2
B P
U T G
D R I L L I N G
T E C H N O L O G Y
P R O J E C T
Figure B-7 illustrates a good WOB record. Note that the recommended profile uses an assumed soil shear strength at the site, which explains the slight difference between the predicted and actual WOB profiles before reciprocation starts.
Figure B-7–An Example of a Good Conductor Jetting WOB Record and Comparison with the Recommended Profile
B-14
DWGOM-GP 02-1019
Draft D, 12/13/02
2 0 0 2
B P
U T G
D R I L L I N G
T E C H N O L O G Y
P R O J E C T
Figure B-8 illustrates poor WOB profiles during jetting. Either the WOB is maximized too soon, in which case no friction is gained from the lower part of the soil profile, or the WOB is not maximized at all, which means that the immediate capacity calculated using the described method grossly overestimates the actual immediate capacity. Although both conductors were installed to target depth, their load bearing capacity has been reduced from what is calculated by the methods presented here. Recommended WOB (kips) 0
0
20
40
60
80
100
120
140
Max available WOB
50
Depth (ft)
100
150
200
250 Figure B-8–Examples of Poor WOB Profiles
B.2.5 RECIPROCATIONS The purpose of reciprocations is to remove friction along the outside of the conductor to allow penetration. Current weights of the conductor and BHA are not sufficient for full penetration without reducing friction. The amount of friction reduction is, at least in part, a function of the amount and stroke length of reciprocation.
Draft D, 12/11/02
DWGOM-GP 02-1019
B-15
2 0 0 2
B P
U T G
D R I L L I N G
T E C H N O L O G Y
P R O J E C T
Excessive WOB Conductor may stick or angle may build
ed sir De B WO
ge ra n
Depth below mudline
WOB
Excessive soil damage Conductor capacity effected
JETCAP WOB profile Figure B-9–Desired and Undesired WOB
It is desirable to reciprocate such that some of the friction is reduced without causing extreme damage to the soil, while at the same time ensuring that the conductor does not get stuck or build an unacceptable angle before the desired penetration is met. This is illustrated qualitatively in Figure B-9. The red area to the left of the recommended WOB profile represents a condition where the soil will be extremely damaged, which will result in slow soil healing times (reduced Da values) and perhaps a permanent loss of soil strength. The red area to the right of the WOB profile represents conditions where the conductor may be unable to be pulled free or excessive
B-16
DWGOM-GP 02-1019
Draft D, 12/13/02
2 0 0 2
B P
U T G
D R I L L I N G
T E C H N O L O G Y
P R O J E C T
angles from the vertical may build in the conductor. The desired WOB profile should fall between these two red regions as shown in Figure B-9. Work performed by the Norwegian Geotechnical Institute suggests that GOM soils with a ratio of the undrained shear strength to the effective vertical stress less than one-half (Su/sv′<0.5) will tend to squeeze into an open hole. This implies that soil would be scraped away on the downward reciprocation stroke causing a reduction in lateral stresses and inducing shear stresses in the soil at points away from the conductor-soil interface. A reduction in lateral stresses will impact the immediate capacity of the conductor. Furthermore, the induced shear stresses will cause pore water pressures to increase and will result in longer healing (setup) times. Since reciprocation is necessary for conductor installation, it is recommended that the length of the reciprocation stroke be minimized and kept less than 10 ft, measured at the mudline (10 ft up and 10 ft down). By minimizing the length of the open hole, it is possible that the stresses in the soil may arch around the open hole, thus minimizing any squeezing of the soil into the open hole.
Draft D, 12/11/02
DWGOM-GP 02-1019
B-17
2 0 0 2
B.3
B P
U T G
D R I L L I N G
T E C H N O L O G Y
P R O J E C T
RECOMMENDATIONS
B.3.1 DESIGN PROCEDURE The following recommendations summarize the design methodology: 1.
If available, use site-specific soil boring data. If soil boring data is not available, review high-resolution seismic data over the well area. If data do not show signs of active or relic geohazards such as faults, landslides, buried channel, or erosion, capable of influencing the soil shear strength in the top 300 ft, then use the average GOM soil properties given in Table B-2. If active geological processes are suggested, an experienced geotechnical engineer should modify the soil profile. A soil boring may be required for extreme cases.
2.
Calculate the weight to be landed using Eq. B.6. The density of the fluid in the open hole should be carefully considered to take full advantage of the buoyancy provided by the mud. However, consider designing for the load case during the cement job when the lighter weight spacer fluid is in the open hole and the cement is in the inner cement string.
3.
Calculate the required conductor setting depth using Eq. B.11. A minimum safety factor of 1.3 is recommended (i.e., FS1=1.0 and FS2=1.3).
4.
Using Eqs. B.12, B.13, and B.14, construct a WOB profile to be followed during jetting.
B.3.2 DOCUMENTATION It is recommended to carefully document each conductor installation. The following list describes the minimum data that should be recorded.
B-18
1.
A detailed makeup of the conductor and the first string to be landed.
2.
WOB profile from mudline to final depth, including pump volume and rates, reciprocation length and intervals.
3.
Setup time between end of jetting and landing of the first string.
4.
Mud weight in the hole prior to landing, with indication of depths (if any) where the well might be flowing.
5.
Actual weight landed as measured on the hook.
DWGOM-GP 02-1019
Draft D, 12/13/02
2 0 0 2
B P
U T G
D R I L L I N G
T E C H N O L O G Y
P R O J E C T
B.3.3 OPERATIONS Based on the available data, there appears to be no single aspect of the conductor jetting process that clearly causes a conductor to subside. However, back analysis of available data trends, coupled with conventional soil mechanics principles, suggests the following recommendations. Each recommendation listed below alone may not result in conductors that do not subside. However, it is expected that small changes in design and operations will make the difference between success and failure. 1.
Follow the recommended WOB profile. Of the seven conductors that have failed in 2002, none have followed the recommended WOB profile. A drill string with higher over pull capability should allow for more aggressive jetting. Contractors involved in the operational aspects of the jetting procedure indicated the last 100 ft of jetting is difficult, and they are fearful of sticking the conductor. Because of the fear of sticking the pipe, the conductor is jetted in using excessive reciprocation. Jetting the conductor with a drill string capable of providing more over pull may alleviate the problem of getting the conductor stuck.
2.
Minimize the stroke length of the reciprocation to 10 ft or less. It is more desirable to have several 10-ft strokes as compared to one 30-ft stroke.
3.
Minimize the frequency of reciprocations. Maintain a relatively high WOB after reciprocation (70 to 80% of the WOB prior to reciprocating). Do not remove all friction. This will cause less damage and should result in more predictable setup times.
4.
The average GOM shear strength data given in Table B-2 should be used with caution. In fact, the data in Table B-2 should only be used if site-specific soils data is not available and geophysical information does not indicate anomalous features.
5.
Cement the surface casing prior to landing the full weight of the surface casing on the conductor. This will reduce the fully landed weight by about 25% due to the extra buoyancy provided by the cement.
6.
Jet nozzles should be placed 12” to 18” above the conductor shoe.
7.
Land an intermediate liner before landing the surface casing.
8.
Minimize the pump rate to less than 1,000 gpm and pressures below 2,000 psi.
9.
Jet the conductor as fast as reasonably possible.
10. Assume the smallest conductor diameter when running the calculations (e.g., use a 36-in. diameter when determining the jetted length for a 36-in. x 38-in. conductor). 11. Following the recommended WOB profile can result in compressive loads in the BHA above the mudline. There has been some concern that this may build angle in the conductor as the jetting process continues, but three recent jetting operations have been performed, all following and some even exceeding the recommended WOB profile. All three conductors were successfully jetted within acceptable vertical limits.
Draft D, 12/11/02
DWGOM-GP 02-1019
B-19
2 0 0 2
B P
U T G
D R I L L I N G
T E C H N O L O G Y
P R O J E C T
B.3.4 USING JETCAP JETCAP (Jeanjean et al, 1999) is an Excel spreadsheet program developed within UTG for conductor design. JETCAP utilizes Eq. B.6 to calculate the landing load, Eq. B.11 to calculate the conductor length, and Eqs. B.12-B.14 to develop a WOB profile. This design procedure was developed and calibrated using a limited number of case histories. In particular, conductor penetration lengths used to calibrate JETCAP were between 65 and 175 feet, and reciprocation stroke lengths used for JETCAP calibrations were 20 ft or less. Extrapolated results from this limited data set are uncertain. When using JETCAP, the following should be considered: 1.
Use site-specific soil property data if available.
2.
When computing the landing load, consider designing for the peak loading case. If appropriate, use a combination of drilling mud and lighter weight spacer fluids in the open hole and cement inside the inner string. Typically, landing weights are computed assuming 12.0 to 12.5-ppg mud in both the open hole and inner string that results in underestimating the peak landing load by about 15 percent.
3.
Input the conductor diameter as 36-in. even if it is a 38-in. x 36-in. conductor.
For conductors greater than 220 ft in length, UTG should be consulted regarding conductor design and jetting procedures.
B-20
DWGOM-GP 02-1019
Draft D, 12/13/02
2 0 0 2
B.4
B P
U T G
D R I L L I N G
T E C H N O L O G Y
P R O J E C T
WELL FAILURES Despite the lessons learned, seven conductors on BP wells subsided in 2002. These include: 1.
Thunder Horse-three conductors have subsided
2.
Mad Dog-one conductor has subsided
3.
Angola (Block 18)-two conductors have subsided
4.
Holstein-one conductor has subsided
The jetted lengths for these conductors range from 210 ft to 314 ft, with the majority of the jetted lengths greater than 260 ft. In each of these failures, none of the installations followed the recommended WOB profile discussed in section B.2.4.
Draft D, 12/11/02
DWGOM-GP 02-1019
B-21
2 0 0 2
B.5
B P
U T G
D R I L L I N G
T E C H N O L O G Y
P R O J E C T
GLOSSARY OF TERMS BHA
Bottomhole assembly
BOP
Blowout preventer
CADA
Cam-Actuated Drill Ahead tool
JETCAP Design program designed by UTG, used to compute lengths and weights
B-22
POOH
Pull out of hole
ROP
Rate of penetration
ROV
Remotely operated vehicle
WOB
Weight on bit
DWGOM-GP 02-1019
Draft D, 12/13/02
2 0 0 2
B.6
B P
U T G
D R I L L I N G
T E C H N O L O G Y
P R O J E C T
REFERENCES 1.
Beck, R.D.; Jackson, C.W.; and Hamilton, T.K. (1991) “Reliable Deepwater Structural Casing Installation Using Controlled Jetting”.
2.
Frugo Limited.
3.
Frugo-McClelland Marine Geosciences.
4.
Jeanjean, P.; Trevor, T.; and Hampson, K. (1999) “BP Amoco UTG Recommended Practice for Jetted Conductor Setting Depth for Gulf of Mexico Deepwater Applications”, Report No. 9921CEH001.
5.
Jeanjean, P. (2002) “SPE 77357, Innovative Design Method for Deepwater Surface Casings”.
6.
Liedtke, E. (2002) “Memorandum, Assessment of conductor subsidence at Thunder Horse”.
7.
Norwegian Geotechnical Institute (2000) “BP Amoco UTG Recommended Practice for Jetting Conductor Setting Depth for Gulf of Mexico Deepwater Applications”.
Draft D, 12/11/02
DWGOM-GP 02-1019
B-23
2 0 0 2
B P
U T G
D R I L L I N G
T E C H N O L O G Y
P R O J E C T
APPENDIX A DERIVATION OF THE REQUIRED CONDUCTOR LENGTH, L The capacity of the jetted conductor can be expressed as: Qt = Q0 + SETUP = WOBlast + 0.055[2 + log(t)](πDL)(SuAVE)
(A.1)
with: t = Time, in days Qt = Conductor capacity at time = t WOBlast = Weight on bit measured at final penetration D = Conductor diameter L = Conductor length below mudline SuAVE = Average undrained shear strength over the length of the conductor The last WOB measured is assumed to be proportional to the available weight during jetting. The constant of proportionality, called the WOB utilization ratio, is noted R. As discussed in the main text, R should be close to 0.8 as per past practice. It should not exceed 1.0 in order to maintain tension in the jetting assembly. Therefore: Q0 = WOBlast = R(Wcond + WWH + WDC + WCADA)
(A.2)
with: Q0 = Conductor capacity immediately after jetting (t = 0.01 days) WOBlast = Last weight on bit recorded during installation R = WOB utilization ratio, should be equal to 0.8 Wcond = Buoyed weight of the conductor, in water
B-24
DWGOM-GP 02-1019
Draft D, 12/13/02
2 0 0 2
B P
U T G
D R I L L I N G
T E C H N O L O G Y
P R O J E C T
WWH = Buoyed weight of the wellhead housing, in water WDC = Buoyed weight of the drill collars, in water WCADA = Buoyed weight of the CADA tool, in water The weight of the conductor, Wcond, and the weight of the drill collars, WDC , are functions of the unknown conductor length L. In order to resist bending moments close to the mudline, the conductor usually has a thicker wall section in the upper 80 ft close to the mudline. Therefore: Wcond = w1L1 + w2(L – L1)
(A.3)
with: w1 = Weight per unit length of upper conductor section in water L1 = Length of upper conductor section (usually equal to 80 ft) w2 = Weight per unit length of lower conductor section in water WDC = wDCL
(A.4)
with: wDC = Weight per unit length of drill collars in water Combining A.3 and A.4 into A.2 gives: Q0 = R[w1L1 + w2(L - L1) + WWH + wDCL + WCADA]
(A.5)
The average shear strength along the conductor length L can be expressed as: SuAVE = Su0 + Su1L
(A.6)
with: Su0 equals 0.0191 ksf for an average GOM soil profile Su1 equals 0.0043 ksf/ft for an average GOM soil profile Combining A.5 and A.6 into A.1 gives the total capacity: Qt = R[w1L1 + w2(L - L1) + WWH + wDCL + WCADA] + 0.055[2 + log(t)](πDL)(Su0 + Su1L)
(A.7)
On the other hand, the resisted load can be expressed as: LOAD = FS1(Wcond + WWH) + FS2(Wlanded)
Draft D, 12/11/02
DWGOM-GP 02-1019
(A.8)
B-25
2 0 0 2
B P
U T G
D R I L L I N G
T E C H N O L O G Y
P R O J E C T
with: LOAD = Load to be resisted by the conductor Wlanded = Load to be landed on conductor (do not include self-weight of conductor) FS1 = Safety factor on weight of conductor and wellhead FS2 = Safety factor on weight to be landed If the quantities Wlanded, FS1, FS2 are given, the load to be resisted is also a function of one unknown, L. Since the load and resistance are equal: Qt = LOAD
(A.9)
Substituting Eqs. A.7 and A.8 into A.9 and re-arranging yields the following quadratic equation: AL2 + BL + C - (FS2)(Wlanded) = 0
(A.10)
with: L = Required conductor length below mudline A = 0.055[2 + log(t)](pD)(Su1) B = R(wDC + w2) + 0.055[2 + log(t)](pD)(Su0) - FS1w2 C = R[WWH + WCADA + L1(w1 – w2)] - (FS1)(WWH) + (FS1)[L1(w2 - w1)] The solution to Eq. A.10 can be expressed as follows, all quantities as previously defined:
L=
- B + B 2 - 4 AC + (4 A)( FS2 )(Wlanded ) 2A
(A.11)
The required conductor length below mudline is a function of the square root of the weight to be landed on the conductor.
B-26
DWGOM-GP 02-1019
Draft D, 12/13/02
2 0 0 2
B P
U T G
D R I L L I N G
T E C H N O L O G Y
P R O J E C T
APPENDIX B EXAMPLE CALCULATION: LANDING LOAD Given the following data, calculate the landing load. Water Depth: 3,775 ft Drill pipe cementing/landing string OD: 5-in. Drill pipe cementing/landing string ID: 4.276-in. Drill pipe actual weight: 22.6 lbf/ft Surface casing:
22 (wall=1.0-in.) 224 3,775 - 5,000 ft 22 (wall=0.625-in.) 143 5,000 - 6,275 ft
12.5 ppg mud inside drill pipe 12.5 ppg mud inside open hole below mudline 8.6 ppg seawater occupying drill pipe x surface casing annulus Solution (Volume Displacement method):
Wlanded = Wsurf + Wcmt + Wcmt-fld + Wcmt-surf-fld - Wsurf-fld Wsurf = (224)(5,000 - 3,775) + (143)(6,275 - 5,000) = 456,700 lbf Wcmt = (22.6)(6,275-3,775) = 56,500 lbf Vcmt-fld = .7854(4.276/12)2(2500)(7.48/42) = 44.4 bbl Wcmt-fld = (44.4)(12.5)(42) = 23,300 lbf Vcmt-surf-fld = 0.7854(20/12)2(1,225)(7.48/42) + 0.7854(20.75/12)2(1,275)(7.48/42) -0.7854(5/12)2(2,500)(7.48/42) = 948.5 bbl
Draft D, 12/11/02
DWGOM-GP 02-1019
B-27
2 0 0 2
B P
U T G
D R I L L I N G
T E C H N O L O G Y
P R O J E C T
Wcmt-surf-fld = (948.5)(8.6)(42) = 342,600 lbf Vsurf-fld = 0.7854(22/12)2(2500)(7.48/42) = 1175.3 bbl Wsurf-fld = (1175.3)(12.5)(42) = 617,000 lbf Wlanded = 456,700 lbf + 56,500 lbf +23,300 lbf +342,600 lbf – 617,000 lbf = 262,100 lbf
S OLUTION (P RESSURE -A REA
METHOD ):
Wlanded = PtAt - PbAb - PxoAxo + Wsurf + PziAzi - PzoAzo Pt = 0.052(8.6)(3,755) = 1,688 psi At = 0.7854(222 - 52) = 360.5 in2 Pb = 0.052(12.5)(6,275) - 0.052(8.6)(6,275 - 3,775) = 2,961 psi Ab = 0.7854(202 - 52) = 294.5 in2 Pxo = 0.052(12.5)(6,275) - 0.052(8.6)(6,275 - 5,000) = 3,509 psi Axo = 0.7854(20.752 - 202) = 24.0 in2 Wsurf = (224)(5,000 - 3,775) + (143)(6,275 - 5,000) = 456,700 lbf Pzi = 0.052(12.5)(6,275) = 4,079 psi Azi = 0.7854(20.752) = 338.2 in2 Pzo = 0.052(8.6)(3,775) + 0.052(12.5)(6,275 - 3,775) = 3,313 psi Azo = 0.7854(222) = 380.1 in2 Wlanded =
(1,688)(360.5) - (2,961)(294.5) - (3,509)(24.0) + (456,700) + (4,079)(338.2) -(3,313)(380.1) = 229,200 lbf
The difference between the two solutions is due to the formulations used by both methods. The volume displacement method was formulated to take into account the weight of the cementing string, whereas the pressure area method was not. Either method can be reformulated to consider the most appropriate condition.
B-28
DWGOM-GP 02-1019
Draft D, 12/13/02
2 0 0 2
B P
U T G
D R I L L I N G
T E C H N O L O G Y
P R O J E C T
APPENDIX C EXAMPLE CALCULATION: REQUIRED CONDUCTOR LENGTH Given the following conductor data below, calculate the required conductor length to accommodate an estimated 230 kip landing load with 3 days of healing time. Use a safety factor of 1.5 on the weight to be landed, and a utilization ratio of 0.8. CADA tool = 3,500 lbf (in air) Wellhead housing = 10,000 lbf (in air) Drill collars = 209 ppf (in air) Water Depth: 3,775 ft Conductor casing:
36 (wall = 2.0-in.) 727 ppf (80 ft) 36 (wall = 1.0-in.) 374 ppf
Su0 = 0.0191 ksf Su1 = 0.0043 ksf/ft
S OLUTION : A = 0.055[2 + log(t)](pD)(Su1) A = 0.055[2 + log(3)](p)(3 ft)(4.3 lbf/ft3) = 5.52 lbf/ft2 B = R(wDC + w2) + 0.055[2 + log(t)](pD)(Su0) - FS1w2
(
wDC = 1-
(
w2 = 1-
Draft D, 12/11/02
)
8.6 209 = 182 ppf 65.44
)
8.6 374 = 325 ppf 65.44
DWGOM-GP 02-1019
B-29
2 0 0 2
B P
U T G
D R I L L I N G
T E C H N O L O G Y
P R O J E C T
B = 0.8(182 + 325)lbf/ft + 0.055[2 + log(3)](p)(3 ft)(19.1 lbf/ft2) - (1.0)(325 lbf/ft) B = 105.1 lbf/ft C = R[WWH + WCADA + L1(w1 – w2)] – (FS1)(WWH) + (FS1)[L1(w2 – w1)]
(
)
wWH = 1-
8.6 10, 000 = 8, 686 lbf 65.44
(
8.6 3,500 = 3, 040 lbf 65.44
wCADA = 1-
(
w1 = 1-
)
)
8.6 727 = 632 lbf/ft 65.44
C = 0.8[8,868 lbf + 3,040 lbf + 80 ft(632 – 325)lbf/ft] - (1.0)(8,686 lbf) + (1.0)(80 ft)(325 - 632)lbf/ft = -4.217 lbf
L=
L=
B-30
- B + B 2 - 4 AC + (4 A)( FS2 )(Wlanded ) 2A 2 -105.1+ (105.1) - 4 (5.52) (-4, 217 ) + 4 (5.52)(1.5) ( 230, 000)
2 (5.52)
DWGOM-GP 02-1019
= 242 ft
Draft D, 12/13/02
