Vibration Assessment of Complex Pipework

Transcription

1 Vibration Assessment of Complex Pipework DNV GL Technology Week Aravind Nair 31 Oct SAFER, SMARTER, GREENER

2 Overview Vibration Induced Fatigue- Sources, Consequence; State of the art-pipeline VIV Complex Pipework- Jumper System VIV assessment challenges Jumper VIV JIP Future study

3 Vibration Induced Fatigue Main sources: Internal Flow Induced Vibration Common sources- major discontinuities such as valves, pipe bends, intersections etc. External Flow Induced Vibration Vortex induced vibration -shedding of vortices downstream of the pipe system Flow past dead legs Mechanical Excitation Presence of pumps/ compressors, chokes etc Internal and External Flow Induced & Acoustic Vibration Identify source of vibration Frequency and consequence of vibration Failure & Mitigation Study Ensure operational integrity Even this can result in Vibration!

4 Failures PHMSA bulletin: ADB 2015 Missouri river Montana Cook Inlet Span Offshore Jumper Failure Mobilisation for repair several million US$ Offshore time million US$/day Loss of income million US$/day Cost may be huge! 4

5 What is VIV and state of the art? Current Cross-flow (Lift) f shedding = St U D In-line (Drag) Image courtesy: Cesareo de La Rosa Siqueira ( f shedding = vortex shedding frequency St = Strouhal number (0.17 to 0.25) U = current speed D = diameter Non-linear fluid structure interaction phenomenon Cross-flow vibration amplitudes of the order of one diameter Extensively studied for risers/pipeline free spans Design guidelines and tools available for risers/pipeline free spans Relatively less knowledge/studies for Complex pipe systems 5

6 Uneven seabed - Subsea Reinertsen Engineering

7 What has been done

8 State of the Art on VIV Analysis Response Based Models Force Based Models Flow Based Models Empirical Models (FatFree): Cross and Inline Vibration Sheared Flow Force Based Models (Shear7): Cross Vibration; Sheared flow Inline Vibration Flow Based Models (CFD): Promising Report from the vortex induced vibration, specialist committee of the 25th ITTC (2008)

9 Jumper VIV Challenges 9

10 Jumper Systems Subsea Jumper: short connection- transports fluid between two subsea components Connects Pipeline end terminations. Typical length: >100ft Complex: anything that potentially deviates from DNV RP F105. Shapes: U, M, Z, V and any such complex shapes; manifold piping

11 Jumper VIV: Common Industry Practice Extend existing VIV concepts and methodology for risers/pipelines to jumpers (e.g. use DNV-RP-F105) Use engineering judgment, exercise caution and understand limitations when using DNV-RP-F105 for subsea jumpers DNV-RP-F105 Response Models

12 Challenges-Category CHALLENGES STRUCTURE FLUID INTERACTION DAMAGE ESTIMATION

13 Studied Shapes

14 Effect of Geometry- Modal Response Frequency Differences Mode Shape Differences

15 Modal Response

16 Modal Response

17 Modal Response- VIV Damage There is no empirical data to validate any of the damage assessment calculations on non-straight geometries. Comparison of fatigue damage due to the modal differences Only a 1st step The calculated damage assumes all other parameters (such as vortex shedding patterns), are similar to straight pipelines and that the currently available response model (DNV RP F105) is valid.

18 Fluid Interaction: Available Studies ExxonMobil s Model test- M Shaped Jumper Wang et.al, VIV Response of a Subsea Jumper in Uniform Current, OMAE st and ONLY model test data available for jumper systems. CFD Study Holmes, S (Red Wing Engineering) and Constantinides, Y. (Chevron Energy Technology Company); Vortex Induced Vibration Analysis of a Complex Subsea Jumper, OMAE DNV Study- CFD: M shaped Jumper form DNV s ISOPE 2013 paper M shaped Jumper from ExxonMobil s OMAE 2013 paper

19 Damage Estimation: Available Studies There are few studies available for damage estimation. Studies typically use parameters in DNV RP F105 (Free span pipelines) for VIV damage estimation Limited studies account for unique factors affecting jumper systems No study is Validated Studies: Nair, A, Sharma, P, Grytoyr, G., Fyrileiv, O, and Vedeld, K (2013), VIV Assessment of Rigid Jumper Systems-A Comparative Study on Jumper Shapes, 23rd International Offshore and Polar Engineering Conference ( ISOPE 2013), June 30-July 5, 2013, Anchorage, USA Nair, A, Kadiyala, R and Whooley, A (2010). Vortex Induced Vibrational Assessment of Multi-planar Geometry-New Methodology, Proc. Of the Deep Offshore Technology Conf., Amsterdam, The Netherlands, DOT 2010, PennWell Corp. Vedeld, K., Sollund, H., Fyrileiv, O. (2011), Fatigue and Environmental Loading of Large Bore Manifold Piping, Proc. of the ASME th International Conference on Ocean, Offshore and Artic Engineering, OMAE2011, OMAE Vedeld, K., Sollund, H., Fyrileiv, O. (2011), Fatigue and Environmental Loading of Large Bore Manifold Piping, Proc. of the ASME th International Conference on Ocean, Offshore and Artic Engineering, OMAE2011, OMAE Hariharan, M, Cerkovnik, M (2H Offshore Inc.), Thomson, H. (Chevron Texaco), The Significance of Low Velocity Near Bottom Currents on the In-Line Vortex-Induced Vibration Response of Rigid Subsea Jumpers

20 The issue Designers discretion and judgement: What am I comfortable with? The Significance of Low Velocity Near Bottom Currents on the In-Line Vortex-Induced Vibration Response of Rigid Subsea Jumpers:, Madhu Hariharan, Mark E. Cerkovnik (2Hoffshore Inc) and Hugh M. Thompson (ChevronTexaco),? VIV Assessment of Rigid Jumper Systems-A comparative study on Jumper shapes, A. Nair, P. Sharma, G. Grytøyr, O. Fyrileiv, K. Vedeld, ISOPE 2013 New Approach to the Design of Rigid Jumpers for Free Standing Hybrid Risers, S Plouzennec, M. Sonawane, T. Eyles,IBP1708-1

21 Damage Estimation Contd.. Geometry/ Client: Undisclosed

22 JIP Study 2

23 JIP Study- Phase I (Completed) Key Deliverable part of the Guidance Report Generate Jumper specific Response Curve Compare against DNV-RP-F105 FEA-CFD Study DNV-RP-F105 Curve

24 JIP to address Jumper VIV started in December 2014 Objective: EITHER update DNV-RP-F105 incorporating Jumper response OR a new standalone recommended practice on Jumper VIV Five participants in phase 1 BP DNV GL ExxonMobil Petrobras Saipem

25 Phase-1: Data Collection and Analysis Wang, H et.al VIV Response of Subsea Jumper in Uniform Current, OMAE 2013 [ExxonMobil Paper] Ay/D plots Exxon Mobil s model test data 10 OD jumper Base case 141 towing experiments Mode shapes

26 Detren d Integr ate Detren d Integr ate Data Assessment Raw Acceleration Velocity Displacement Total RMS Displacement Sample Processing Approach Only- Not JIP data

27 Approach X Y Z

28 New Jumper Specific Response Curve DNV-RP-F105 Response Models Jumper Response Curve currently shows differences with DNV- RP-F105 curve. NEW Guidance Report issued.

29 FSI CFD Study The purpose of the study is to evaluate the applicability of CFD as a response prediction tool in-lieu of model tests. Five cases selected for preliminary comparison against the model test results.

30 Jumper VIV CFD Activity Flow around stationary cylinder (Pre-JIP) 2D Simulations 3D Simulations - RANS vs DES Fluid-Structure Interaction (JIP) Analysis methodology Couple FE Structural solver and CFD solver Solve structural response within CFD code using linear mode superposition (PFSI Practical FSI) Both methodologies used in the JIP CFD Study

31 Stationary Cylinder Study Cases Flow Past 2D Rigid Cylinder at U = 0.3 m/s (CASE1-2D), 0.7 m/s (CASE2-2D),0.9 m/s (CASE3-2D) Calculate quantities such as C D, C L, St with respect to Re. Compare with experimental results D L D (m) 0.06 L (m) 0.3 ( 5D) U (m/s) 0.3, 0.7,0.9 Re (x10 3 ) 18.6, 45.0, 55.5

32 1.2 m ( 20D) 2D Mesh Set Up Mesh Size = 16,111 FV s, x= y = 0.003m (0.05D) around cylinder, y+ < 1 SYMM 2.4 m ( 40D) U INLET OUTLET SYMM

33 Case 1,2 and 3: Comparison of force quantities CASES U (m/s) Re (x10 3 ) C d (C L ) rms St Case Case Case C D = 1.2 for 10 4 <Re<2.8x10 5 St* = 0.2 for 10 4 <Re<10 5 C D = 1.2 DNV RP-C205 *St = f s D /U

34 3D Mesh Set Up t = s x= y= z=0.003m (0.05D) RANS: k- Model DES 5D

35 Flow Around Cylinder - Comparison of Aerodynamic Coefficients (RANS) CASES U (m/s) Re (x10 3 ) C d (C L ) rms St Case 1-2D Case 2-2D Case 3-2D CASES U (m/s) Re (x10 3 ) C d (C L ) rms St Case 1-3D Case 2-3D Case 3-3D

36 JIP FEA-CFD Comparison between AcuSolve PFSI study and Ansys-Fluent Study Similarity and Differences with Experiment Preliminary Guidance for using CFD study

37 Remaining Challenges Effect of current direction; and the differences in the vortex pattern based on current direction and the angle of the structural members (legs). Shielding effect from adjacent legs Hydrodynamic damping Presence of coupled flexure-torsion and shear Fatigue Estimation Response model for range of jumper sizes and shapes The FSI for jumper is more complex than for a freespan Cross-flow Current Freespan In-line Multiple competing excitation regions possible on an M-shaped jumper Cross-flow (horizontal In-line leg) Cross-flow (vertical leg) In-line FSI: Fluid-Structure Interaction Current M-shaped Jumper

38 Phase II JIP Plan Desk Top Study Additional load cases (esp. off angle cases) using FEA-CFD Special topics such as fatigue estimation.

39 Thank You Aravind Nair (970) [Mobile] SAFER, SMARTER, GREENER