Study of Feedback Controlled Variable Cone Expansion Process
Prepared for presentation at the 2013 SIMULIA Community Conference Allan Zhong John Gano
Outline
1. Back Backgr grou ound nd 2. Prel Prelim imin inar arie iess 3. The The FEA FEA mode modell 4. Friction Friction effe effect ct on variable variable cone expansion expansion process 5. Concl Conclusi usion onss and rema remarks rks 6. Ackn Acknow owle ledg dgem emen entt
2/26
Background Why do we expand pipes? Open Hole Completion - Expandable screen for sand control
Liner Hanger - Expandable liner hanger
Ref: SPE 101538 Unique Expandable Sand Screen and Expandable Liner Hanger for Saudi Aramco 3/26
Background (continued) – What is Variable Cone Expansion? It is a process that is realized through a two-cone expansion system with a smaller sized, fixed cone at the front and a variable cone at the back. The variable cone can be moved between an expanded position and a retracted position. During expansion, the variable cone is enlarged to its expanded position and advanced through the casing until a restriction (e.g. smaller ID wellbore outside the casing) is reached.
4/26
Background (continued) – Feedback Control System The movement of the variable cone is controlled by a feedback control system. The hydraulic feedback control process can simply be described as the force applied on the wedge, controlled by the force acted on the cone system.
Assume the load on variable cone is F1, then the load on the wedge F2 is related to the F1 in the following way F2=m F1 when F1 < F_max kips F2= F_min when F_max < F1 5/26
Background (continued) – Why Variable Cone Expansion?
To allow expansion through restrictions without incurring large force The technology also enables robust performance of expandable liner hanger
6/26
Background (continued) – Simulation of Variable Cone Expansion From numerical simulation point of view, a key feature of the expansion process is that the force on wedge is dependent of the force on variable cone. In other words, the load on wedge is feed back controlled. This type load could not be applied in Abaqus until a couple of years ago. To simulate the variable cone expansion, a user defined sensor can be used. Note: This work was performed before a user defined sensor was available in Abaqus, and a user element was developed to achieve the feedback control.
7/26
Objectives of This Study 1) Determination of critical friction coefficient 2) Study the influence of friction coefficients between cone and pipe, and wedge and variable 3) Improve design to prevent self locking
8/26
Preliminary – Critical Friction Coefficient Critical Friction Concept
F cos F sin
c
tan
Note: The critical friction coefficient is a structural property. In this example, the larger the angle θ, the larger the critical friction coefficient is. 9/26
Preliminary – Verification of Feedback Control The verification is performed on a very simple model with two separate rectangular blocks: the force on one block F1, is applied, the controlled force F2 is applied on the other block. The control scheme is: F1= 0 to F_max lb F2= m*F1 if F1< F_max lb F2= F0 if F1 >= F_max lb
10/26
Preliminary – Verification of Feedback Control (continued)
The blue curve is the active load; the brown curve is the controlled force. There is a delay in controlled force magnitude. The controlled force varies as specified per the control scheme except it has a time delayed response. The delay, of course, is expected for a hydraulic controller. It is noted that the delay can be reduced by a decreased time 11/26 increment in the FEA model.
FEA Model for Variable Cone Expansion
Simplified variable cone - shown one branch only 12/26
FEA Model for Variable Cone Expansion (continued) wedge
casing
F1
F2
Fixed cone Variable cone
To simulate the loading conditions efficiently: 1) The variable cone is fixed axially by a rigid surface but can move radially 2) The active load is the force pulling the casing 3) The wedge is pushed under the controlled load 13/26
Critical Friction Coefficient for the Base Design The critical friction coefficient is determined straight forward, the wedge/cone friction coefficient is varied from low to high to see when a variable cannot reach its fully expanded position (i.e. 8” OD) 7.10” 7 deg cone – expansion at different friction
cone/pipe friction coefficient 0.05 Wedge/cone, cone/cone COF =0.2 Variable cone reached fully expanded state
14/26
Critical Friction Coefficient for the Base Design (continued) 7.10” 7 deg cone – expansion at different friction
cone/pipe friction coefficient 0.05 Wedge/cone, cone/cone COF =0.25 Variable cone did not reach fully expanded state - this leads to under expansion
15/26
Critical Friction Coefficient for the Base Design (continued) 0.2
0.15
The higher the friction the slower the variable cone reaches 8” OD. At the critical friction coefficient it will take very long time for the cone to reach 8” For this case critical friction is ~ 0.2 0.25
16/26
Factors that Influence Critical Friction Coefficient 7.10” 7 deg cone - other friction coefficient Wedge/cone 0.2, cone/pipe 0.05
Wedge/cone 0.2, cone/pipe 0.03
Critical wedge/cone friction depends on cone/pipe friction coefficient
17/26
Factors that Influence Critical Friction Coefficient (continued) 7.10” 7 deg cone - load ratio 7.10 OD X 7 DEGREE X 1.869 RATIO
7.10 OD X 7 DEGREE X 1.32 RATIO
μ pipe
μ pipe
0.07 0.05
YES
0.05
0.05
NO
0.10
μ critical
0.07
0.03
YES
0.05
0.03
NO
0.10
yes
0.15
yes
yes
0.15
0.20
yes
no
0.20
0.25
no
μ critical
0.25
yes
0.30
0.30
yes
0.35
0.35
no
0.40
0.40
0.50
0.50
Increase of load ratio from 1.32 to 1.869 leads ~ 100% increase of critical friction coefficient, from ~ 0.15 to ~ 0.3 18/26
Factors that Influence Critical Friction Coefficient (continued) 7.10” 10 deg cone - change fix cone angle Cone/pipe friction 0.03 Critical friction ~ 0.225
Increase fix cone angle increase critical friction 19/26
Critical Friction Coefficient - Design Consideration 7.10” 10 deg cone, higher load ratio
Cone/pipe friction 0.03, cone/wedge friction 0.375
20/26
Critical Friction Coefficient - Design Consideration (continued) 7.10” 10 deg cone, higher load ratio
Cone/pipe friction 0.03, cone/wedge friction 0.375
0.375 is approximately the critical friction coefficient – as before, it takes a long stroke for the variable cone to move to max OD position under critical friction 21/26
Critical Friction Coefficient - Design Consideration (continued) 7.10” 10 deg cone, higher load ratio 7.25 OD X 10 DEG X 1.32 RATIO
μ critical
7.25 OD X 10 DEG X 1.82 RATIO
μ pipe
μ pipe
0.03
0.03
0.05
yes
0.05
0.10
yes
0.10
0.15
yes
0.15
0.20
yes
0.20
0.225
yes
0.25
no
0.30
yes
0.30
no
0.375
yes
0.40
no
0.40
no
0.50
no
μ critical
0.50
0.25
Critical friction coefficient increased to 0.375 22/26
Summary and Concluding Remarks System Integration Test Observations – when variable cone OD reached 8”, pipe OD is 8.75” (8+2*0.375, pipe wall thickness is 0.375”) - the final design met all performance requirements O.D. - JOINTS 1, 2, & 3 9 8.75 8.5 8.25 CONNECTION
SCREENS
CONNECTION
8 7.75 7.5 7.25 7 0
10
20
30
40
50
60
23/26
Summary and Concluding Remarks (continued) 1. The method developed for simulation of feedback controlled load is verified to function as designed. 2. The method is successfully applied to determine critical friction coefficient for two fixed cone designs under different scenarios. 3. The critical friction coefficient between production cone and wedge is between 0.2 and 0.25 for cone/pipe friction 0.05; it is between 0.15 and 0.2 for cone/pipe friction 0.03. 4. The modified fixed cone, 7.25”/10 deg cone, under same conditions, would lead to higher critical friction as expected, ~ 0.225 for cone/pipe friction 0.03 5. An effective way to increase critical friction is to increase wedge force/expansion force ratio. It is shown that under cone/pipe friction 0.03, the critical friction increased from 0.225 to 0.375. 6. A design with implementations of the design changes considered here was tested successfully. 24/26
Summary and Concluding Remarks (continued) Remarks – a. The limitation of the FEA is that how friction is changing during the expansion process has to be assumed – currently constant friction is assumed. Pressure dependent friction coefficient can be included in the FEA model, which will need to develop another user subroutine. b. Vibration of the expansion system during operation can influence the expansion process, which is not accounted for in the FEA model.
25/26
Acknowledgement
The authors are grateful to Halliburton Management for permission to publish this work.
26/26
