Sunteți pe pagina 1din 15

2/24/2010

Deepwater Rigid Spools:


Design Challenges
and Solutions

OPT Amsterdam
February 2010

Philip Cooper
INTECSEA UK

Talk Outline

X Background
• Available guidance
X Functional requirements and construction considerations
• Rigid tie-in spools for deepwater flowline systems
X Selected Design Challenges
• Design of bends
• Slugging fatigue
• Connector misalignment (probabilistic methods)
• Structure settlement
X The Future

2 OPT 2010: Deepwater Rigid Spools

1
2/24/2010

Spools vs. Flowlines: Cost per m

X Flowlines
• $$$$$$$$$$$
X Spools
• $$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$
$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$
$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$
$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$
$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$
$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$
$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$
$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$
$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$
X Installed costs of spools can be comparable to flowlines
• Procurement, Fabrication, Metrology, Transportation, Deployment,
Testing

OPT 2010: Deepwater Rigid Spools

Background

X Deepwater Rigid Spools: key component of flowline system


X Limited design guidance in codes or literature
• Projects
P j t need
d tto “invent
“i t own rules”
l ” for
f many aspects
t off design
d i
X Little consensus among practitioners on design approach
and criteria
X Need to seek common understanding of best practice for
future projects, and ultimately formalise guidance in codes

4 OPT 2010: Deepwater Rigid Spools

2
2/24/2010

Functional Requirements

X Connect adjacent flowline segments (or flowline to riser)


X Absorb relative movements of ends of spool by elastic
deformation
• Flexibility required
• Minimise interface loads
• Acceptable fatigue life
X Incorporate diverless end connections
• Horizontal
• Vertical
X Other considerations
• Thermal, Corrosion etc

5 OPT 2010: Deepwater Rigid Spools

Construction and Installation

X Sufficiently compact to be:


• Fabricated and handled onshore
• Transported to offshore site
• Installed without unduly large vessel requirements
• Capable of recovery/re-deployment during field life
X Incorporate adjustability to allow final make-up to proceed
rapidly following metrology
• Basic spool geometry must be able to adjust to suit variation
associated with target
g boxes of adjoining
j g structures
• Minimum number of “metrology welds”

6 OPT 2010: Deepwater Rigid Spools

3
2/24/2010

Construction and Installation: Reality

OPT 2010: Deepwater Rigid Spools

Selected Design Challenges

X Effects of external pressure


• Spool typically exposed to full external pressure at some point during
pre-commissioning
pre commissioning or operation
• Well established limit state criteria for straight pipe, but
• No limit state criteria for bends
X Effects of slug flow (inevitable in multiphase flowlines)
• Gravity loads
• Impulse loads (momentum change at bends)
X Effects of end connector misalignment
• Deterministic or probabilistic approach?
X Effects of long term settlement of adjoining structures

8 OPT 2010: Deepwater Rigid Spools

4
2/24/2010

Design of Bends

X No guidance for external pressure design of bends


• DNV OS-F101 suggests a safety factor of 3 on external pressure
• Other codes are silent
X Nonlinear FEA used to simulate ultimate capacity of bends
under external pressure, together with mechanical loads:
• Bend opening
• Bend Closing
• Out of plane bending
• Ovality and circumferential thickness variation included
• Elastic-Plastic Material Properties
X Bend opening and OOP bending capacity found to be similar
to equivalent straight pipe
X Bend closing capacity significantly lower
9 OPT 2010: Deepwater Rigid Spools

Ultimate Load Simulation of Bend

1. FE Model (fine mesh, brick elements,


elastic-plastic material) 2. In-Plane Closing Failure Mode
End 1 y

5D

5D
x
o

5D

End 2

4. Moment-Rotation Response
3. In-Plane Opening Failure Mode 600
5.63°,
Out-of-plane Bending Moment (kN.m)

511kN.m
500

400

300

200

100

0
0 2 4 6 8
Bend Leg End Rotation Angle (Degree)

10 OPT 2010: Deepwater Rigid Spools

5
2/24/2010

OPT 2010: Deepwater Rigid Spools

Bends – Tentative Limit State Criteria

X Form of DNV OS-F101 formula for combined load capacity of straight pipe
adopted 2 2
⎧ M Sd ⎧⎪ γ γ S
2
⎫⎪ ⎫ ⎧ Pe − Pmin ⎫
⎪ ⎪
⎨γ m γ SC + ⎨ m SC Sd ⎬ ⎬ + ⎨γ m γ SC ⎬ ≤1
⎪⎩ α c M (
p 2t ) ⎪⎩ α c S p (t 2 ) ⎪⎭ ⎪⎭ ⎩ Pc (t 2 ) ⎭

X Effective axial force close to zero in tie-in spools


X Characteristic moment resistance at selected pressure taken as ultimate load
from FEA
M Sd
γ m γ SC ≤1
α c M p (t 2 )
X Maximum functional moment capacity developed by applying respective DNV
safety factors
M Sd = M F γ F γ c + M E γ E + M I γ F γ c + M A γ A γ c

X Further work needed to generalise approach and develop widely applicable


guidance

12 OPT 2010: Deepwater Rigid Spools

6
2/24/2010

Slug Load Effects

X Multiphase flow in long flowlines often features segregation


of gas and liquid into “slug” and “bubble”
• Typically 10m – 100m long (comparable to length of spool)
• Several million slugs in lifetime: high cycle fatigue
• Impulse loads at bends can be significant (10 m/s not unusual)
X Change in contents density causes obvious (but often
ignored) gravity load effects on spans
• “Floating” spools see severe fatigue damage
• Seabed or other support needed to achieve sufficient life
X Unsupported bends liable to oscillate in response to impulse
loads
X Complex nonlinear dynamic FEA can be used to simulate
these combined effects
13 OPT 2010: Deepwater Rigid Spools

Slug Flow Fatigue Analysis

X Simple hand calculation sufficient to define approximate max


length of span
• Support needed if spool length exceeds limit
• May be structure or seabed
X Rigorous FEA required to confirm acceptable fatigue
performance
L < Lmax
L > Lmax

14 OPT 2010: Deepwater Rigid Spools

7
2/24/2010

Dynamic Slug Fatigue Analysis

1. Schematic of FE Model showing sequenced 2. Stress Time History: Static & Dynamic
impulse loads at bends and moving distributed
gravity load

G(t)

CF1(t) CF2(t)
CFi(t)

3. Frequency Spectrum
FFT_Magnitude
of Stress Response 4. Distribution of Fatigue Damage along Spool
5

4.5

4 Frq=0.039 (T=25s) : Slugging Travel Frequency


35
3.5 (plus harmonics)
FFT Magnitude

2.5

1.5 Frq=0.35 (T=2.85s) : 1st Mode Frequency


1

0.5

0
0.009 0.209 0.409 0.609 0.809 1.009
Frequency

15 OPT 2010: Deepwater Rigid Spools

Connector Misalignment Effects

X Misalignment loads are introduced as a result of


• Subsea Metrology errors
• Fabrication errors
• Stack-up tolerances (if metrology transponder not directly fixed to hub
– typically the case for horizontal connection systems)
X Resulting misalignment load effects can (theoretically)
exceed those associated with pipeline expansion
X Sources of misalignment are random and independent –
infinite combination of load cases possible
X Conventional deterministic (algebraic) combination of worst
possible misalignments is very conservative
X Probabilistic combination basis has been accepted in some
projects
16 OPT 2010: Deepwater Rigid Spools

8
2/24/2010

Misalignment: SIT Examples

X Vertical connectors shown


X Horizontal connector misalignments
can be more severe (stack-up effects)

OPT 2010: Deepwater Rigid Spools

Deterministic v Probabilistic Summation

X Deterministic
• Algebraic Sum (AS)

xAS = xfabrication + xmetrology + xstack − up

X Probabilistic
X Square Root of Sum of Squares (SRSS)

2 2 2
xSRSS = xfabrication + xmetrology + xstack − up

• In above example, reduction = √3/3 = 0.58


(assuming all misalignment components approximately equal)

18
OPT 2010: Deepwater Rigid Spools

9
2/24/2010

SRSS Verification

X Monte Carlo simulation shows that the likelihood to exceed


the bending moments in the spool obtained by SRSS
method is very small:
• The misalignments comprises multiple components (fabrication,
metrology and stack-up)
• The misalignments comprises 5 degrees of freedom at each spool
end (3 translations and 2 rotations)
• Values for each tolerance conform to uniform distribution with mean 0
(zero)
X Rational process needed to rationalise infinite
f number off
misalignment combinations to manageable sub-set for
analysis

OPT 2010: Deepwater Rigid Spools

SRSS Verification Result

X Simplified analytical model used to approximate response of


spool to random misalignment loads

Normal Distribution Bending Moment: Bending Moment: Bending Moment:


Functional Functional + SRSS Functional + AS
Monte Carlo Simulation Result
Probability (-)

Probability of exceedance (<0.001)

-150 -100 -50 0 50 100 150 200 250 300 350 400 450 500 550 600 650
Resulting Bending Moment (kNm)

20
OPT 2010: Deepwater Rigid Spools

10
2/24/2010

Reliability Analysis with Full FEA

X Conventional approach
requires simplified
structural model to be
used in Monte Carlo
simulation (106
simulations)
• Difficult to implement –
simplified models
X New software tools allow
full FEA models to be
used in reliability
analysis, with advanced
sampling techniques
used to estimate
response surface from
small number of FE runs
• Complex behaviour
and multiple
parameters

X Improved Analysis → Reduced Conservatism → Smaller spools


OPT 2010: Deepwater Rigid Spools

Structure Settlement

X Flowline end structures can be substantial items


• May weigh 100 tonne
X Soils typically soft clay
• Compressible
• Long term settlement of structure significant in life of field
X Settlement of structures imparts significant loads on spools
– especially if seabed supported.
• Undue conservatism should be avoided
X C
Conventional
ti l 1D settlement
ttl t calculations
l l ti
• Conservative: resulting spool stresses unacceptable
• Unable to predict rotation effects from eccentric loads
• Not matched by anecdotal evidence from field

OPT 2010: Deepwater Rigid Spools

11
2/24/2010

Typical In-Line Structure

X Golf of Mexico “Sled” shown is rather light


X Trend is for heavier pipeline end structures
• Multiple valves, vents, protection frame etc.
X “Mudmat” Foundation typically used

OPT 2010: Deepwater Rigid Spools

Numerical Modelling of Settlement

X 3D FEA model of foundation and large soil block


• 3D dispersion of stress correctly captured
X Cam Clay constitutive model with porous elasticity
Cam-Clay
• Compressibility included
• Time dependent effects modelled
X Expected eccentric loads applied to foundation
• Vertical, Horizontal, Torsion and Overturning loads in combination
X Progressive settlement of structure modelled
• Much reduced vertical settlement
• Angular settlement response quantified
• Timescale of settlement estimated

OPT 2010: Deepwater Rigid Spools

12
2/24/2010

Settlement Simulation: Evolution of Pore Pressure

OPT 2010: Deepwater Rigid Spools

FEA vs Analytical Settlement Prediction

X Permeable Mudmat (vent holes) lead to rapid settlement


• Difficult to incorporate in analytical model
X FEA indicates generally lower settlement
• 3D stress dispersion accurately captured
180
22yr, 166mm

150

120
ement (mm)

22yr, 96mm

90
22 90
22yr, 90mm
Settle

60
22yr, 49mm

30 1-D Analytical - LB
1-D Analytical - UB
FEA - 30m Depth, Impermeable, FSA Centre
'FEA - 30m Depth, Permeable, FSA Centre'
0
0 5 10 15 20
Time (yr)

OPT 2010: Deepwater Rigid Spools

13
2/24/2010

Way Forward

X Targeted use of advanced analysis methods to reduce


conservatism in spool design
• Improved Analysis → Reduced Conservatism → Smaller spools
• Example techniques outlined above
• Conventional methods and tools continue to play a part in initial
design
X Limit state design
• Load controlled
• Displacement
p controlled
X Reliability approaches
• Full nonlinear FEA used for probabilistic design for many years in
other industries

OPT 2010: Deepwater Rigid Spools

Limit State Design

X Current practice is Allowable Stress Design


• Generally conservative, except….
• External pressure effects not accounted for properly
X Well established load controlled and displacement controlled limit state
criteria for straight pipe, with internal and external pressure
• e.g. DNV OS-F101
X Tentative load controlled criteria proposed for bends
• Further work needed to validate and generalise
• Potential to develop displacement controlled criteria in same way
X Spools are essentially displacement controlled
• Pipeline expansion, riser motion
• Misalignment & structure settlement
X Strain based design of spools would allow significant reductions of spool
size
• Strength and fatigue criteria comparable to flowline lateral buckling?
• Build on existing SAFEBUCK and HOTPIPE knowledge?

OPT 2010: Deepwater Rigid Spools

14
2/24/2010

Reliability Based Design

X ISO, DNV guidance based on target probability of failure


• Partial safety factors calibrated using probabilistic approach
X Spool design governed by multiple uncertain parameters (in
addition to those which apply to flowlines):
• Position of structures within target box (long, nominal or short spool)
• Pipeline expansion/riser motion
• Misalignments
X Comprehensive probabilistic treatment of uncertain spool
loads and resistance expected to permit more efficient
design compared to conventional deterministic approach
• Full nonlinear FEA model used for reliability analysis

OPT 2010: Deepwater Rigid Spools

Acknowledgements

X Complex FEA performed by my colleagues in INTECSEA


UK Advanced Analysis group:
• Tao Zhao
Zhao, Shulong Liu
Liu, Dan Lee
Lee, Sherif Gebaly
X Project photos provided by INTECSEA colleagues around
the world:
• UK, Houston, Perth, Delft, Rio, KL, Singapore

OPT 2010: Deepwater Rigid Spools

15

S-ar putea să vă placă și