overview of design codes for offshore fixed structures · 2017. 11. 20. · outline part 1 – api...
TRANSCRIPT
Overview of Design Codes for Offshore Fixed Structures
Presentation at 離岸風機研討會
Albert Ku, Dec/5th/2017, Taipei, Taiwan
OutlinePart 1 – API RP2A
–Background of API RP 2A, Recommended Practice for Planning, Designing and Constructing Fixed Offshore Platforms
–Historical work in the development of API RP 2A-LRFD 1st
edition (1993)–Development of API RP 2A-LRFD 2nd edition
• Progress/status of API RP 2A-LRFD 2nd edition• API Task Group 19 latest work/AKADEME project• Technical Alignment of API/ISO
Part 2 – API ductility requirement and seismic design provisions
History of API RP 2AOTC 20831 (2010)
Status of RP 2 Series and ISO Alignment
LRFDtask-group
Younan (2014)
Useful Background Papers1. A Series of Late 1980’s Papers on API 2A-LRFD 1st Edition2. API Offshore Structure Standards: Changing Times, OTC 2008, D. Wisch, A.
Mangiavacchi3. RP 2GEO: The New API Recommended Practice for Geotechnical Engineering,
OTC 2010, P. Jeanjean4. API Offshore Standards – Underlying Risk Values and Correlation with ISO,
OTC 2012, D. Wisch, H. Banon, D. Knoll, S. Verret
5. Development of API RP-2A LRFD 2nd Edition, Offshore Structural Reliability Conference 2014, A. Ku, F. Zwerneman
6. Background to New API Fatigue Provisions, OTC 2010, P. Marshall, J. Bucknell
7. New API RP2A Tubular Joint Strength Design Provisions, OTC 2010, D. Pecknold
8. API RP 2EQ – Seismic Design Procedure & Criteria for Offshore Structures, OTC 2010, A. Younan, F. Puskar
9. ISO 19902 Tubular Members Including Damaged and Grouted Members, OMAE 2011, P. Frieze
10. LRFD Calibration of Load Factors for Extreme Storm Loading in Malaysian Waters, Journal of Marine Engineering & Technology, 2014, N. Nicols, R. Khan
Background between API RP-2A LRFD and ISO 19902
API RP 2A-LRFD 1st edition (1993) Used worldwide but not in the US ISO 19902:2007 largely based on API RP 2A-LRFD
1st edition No maintenance of API RP 2A-LRFD between
1993 and 2012 API RP 2A-LRFD 1st edition retracted in 2012
due to lack of maintenance Current effort focuses on adopting ISO 19902
back to US practice (API 2A-LRFD 2nd edition)
Mapping of RP 2 Series and ISO
RP 2A 2nd
LRFD ISO 19902RP 2A-LRFDISO 19902
OTC 20831 (2010), except red-texted boxes
ISO 19901-9
Co-existing WSD 22nd and LRFD 2nd Edition
OTC 20831 (2010)
Co-existing of WSD and LRFDIn the foreseeable future
Correspondence between and Pf
)( fP Is Gaussian probability distribution function
API RP 2A-LRFD 1st Edition Calibration Methodology
Every designed structural member (beam, column, brace, etc.) has a probability of failure (pf). This pfcan also be expressed as reliability index ()
By carefully selecting load and resistance factors it is possible to achieve
Averaged to be similar as implied in WSD
Minimized spread of
By applying simple procedure, can be calculated for various failure modes
Observe from figure to the right, the “averaged” for each of the 2 curves are similar, but spread of LRFD curve is smaller
OTC 5699 (1988)
Comparison of of Different Failure Modes and between WSD/LRFD
Notes:1) Averages taken over important range of W/G
of 2 to 40 except for piles, where W/G is 0.6 to 2. G = L + D and L = 3D
2) by advanced FOSM, Ref: Moses, 81-22
3) Design FormulasWSD: Rn > SF (L + D)
4/3 Rn > SF (L + D + W)where SF and Rn from API RP2A, 12th edition
LRFD: Rn > 1.3D + 1.5LRn > 1.1D + 1.1L + 1. 35W
where Rn is formula in (81-22)OTC 5699 (1988)
Load Factors in API LRFD & ISO 19902 vs. WSD
WSD LRFD
Operating Condition
0.6R = D + L + WR = 1.67D + 1.67L + 1.67W (normalized)
0.95R = 1.3D + 1.5L + 1.22WR = 1.37D + 1.58L + 1.28W (normalized)
Storm Condition
0.8R = D + L + WR = 1.25D + 1.25L + 1.25W (normalized)
0.95R = 1.1D + 1.1L + 1.35WR = 1.16D + 1.16L + 1.42W (normalized)
Note:-R denotes resistance or structural capacity.-D denotes dead loads, L denotes live loads, and W denotes environmental loads due to wind, wave and currents.-Assumes nominal resistance factor of 0.95 for LRFD.
OTC 5882
Member UC Comparison - Platform A Member UC Comparison - Platform B
Member UC Comparison - Platform C Joint UC Comparison - Platform C
Curve on top means more conservative
WSDLRFD
1980’s Case Studies
Adoption of ISO 19902 for API RP 2A-LRFD 2nd
Edition with Following Modifications
Specific target reliability numbers in the Informative annex removed
Design criteria tied to API RP 2MET and robustness level ultimate strength analysis criteria for Gulf of Mexico (GoM)
Tubular Joint Checks Joint check equations aligned with WSD 22nd edition
Replaced ISO minimum joint requirement with API 50% capacity requirement
Foundation section tied in with API RP 2GEO and resistance factors provided for both pile and shallow foundations
Tied in with API RP 2SIM and API RP 2MOP
Other modifications related to API/ISO technical alignments
AKADEME Project
Key Contractors: API Keystone Atkins Digre Energo McDermott (OFD Engineering) Experia
Funded by API to prepare the 2nd edition of RP 2A-LRFD and to assess the consistency of member and joint unity checks (UCs) based on API RP 2A-WSD 21st edition and ISO 19902 using case study platforms
Undertaken by a group of contractors with API Task Group 19 supervision
Voluntary works performed by McDermott and DNV (BSEE project) also included
Case Study PlatformsNo. Description Water
DepthLocation Exposure
CategoryAnalysis Company
1 4-Leg 274 ft Confidential - McDermott/OFD(volunteer work)
2 8-Leg 360 ft Confidential - DNV(BSEE project)
3 Caisson 45 ft GoM L-3 Energo
4 Tripod 360 ft WesternGoM
L-2 Atkins
5 4-Leg 300 ft Central GoM L-1 McDermott/OFD
6 8-Leg 275 ft WesternGoM
L-2 Keystone
Case Study Platform 3
Platform 3 Member and Joint UC Comparisons
Case Study Platform 4
Platform 4 Member and Joint UC Comparisons
Effect of Hydrostatic Pressure on Member UCs -Platform 4
(a) With Hydrostatic pressure (b) Without Hydrostatic Pressure
Case Study Platform 6
Platform 6 Member and Joint UC Comparisons
Effect of Hydrostatic Pressure on Member UCs - Platform 6
(a) With Hydrostatic pressure (b) Without Hydrostatic Pressure
Discussions of Case Study Results Current case study results show higher consistency between WSD and
LRFD member UCs than 1980’s case studies
Overall, member and joint UCs from ISO 19902 and API RP 2A-WSD are consistent. Minimum joint strength requirements are not included in this comparison
Member and joint UCs from ISO 19902 are slightly higher than those from API RP 2A-WSD for platforms dominated by environmental loading (vs. gravity loading)
Member UC comparisons between ISO and API show more scatter with hydrostatic pressure than excluding hydrostatic pressure
New ISO 19902 Proposed Changes
Paul Frieze (PAFA Engineering) investigated Fred Moses’ earlier work and proposed: Gravity load factor reduced from 1.1 to 1.0 when combined with
environmental loads
Partial resistance factor for compression reduced from 1.18 to 1.10
AKADEME Study Platform 4 L-2 structure
3 leg , 3 pile platform
Western GoM
Pile penetration = 220 ft (B & C), 265 ft (A)
Pile diameter = 48 inches
2 conductors (1 inside pile A)
Normally consolidated to slightly overconsolidatedmarine clays
Metocean Parameters
Wave Height = 63 ft
Wave Period = 12.4 seconds
Surface Current Speed = 1.8 knots
Bottom of Current Profile = 200 ft
Wind Speed (1 hr@10m) = 70 knots
Tide and Surge = 2.5 ftAPI AKADEME Proposed ISO Load and Resistance Factors 27
Platform 4 – All Jacket MembersUC Comparison (Storm, 0 Degree)
API AKADEME Proposed ISO Load and Resistance Factors 28
Original Partial FactorsRevised Gravity Load Factor Only
R = ISO/APIμR = 1.015σR = 0.149COVR = 15%N = 239
R = ISO/APIμR = 1.002σR = 0.153COVR = 15%N = 239
R = ISO/APIμR = 0.998σR = 0.144COVR = 14%N = 239
Revised Compression Resistance Factor Only
Revised Load & Resistance Factors
R = ISO/APIμR = 0.987σR = 0.149COVR = 15%N = 239
Please note:μR = Mean ValueσR = Standard DeviationCOVR = Coefficient of VariationN = Sample Size
ISO/API Technical Alignments
Hydrostatic Checks
Conical Transition
Pile Sleeve Grout Equation
Energo Engineering, “AKADEME Project – API RP 2A vs. ISO 19902 Member and Joint UC Comparisons”, API AKADEME Project Report, 2016.
Axial Compression with Hydrostatic Pressure
API WSD (4) = ISO (1)
API WSD (3) = ISO (2)
API WSD (2) = ISO (3)
API WSD (5) = ISO (4)
API vs. ISO Hydrostatic Code Calibration: Axial Compression and Hydrostatic Pressure
Axial Tension with Hydrostatic Pressure
API vs ISO Hydrostatic Code Calibration: Elastic Local Buckling under Hydrostatic Pressure
API vs ISO Hydrostatic Code Calibration: Inelastic Local Buckling under Hydrostatic Pressure
ISO 19902 Combined Axial and Capped-end Stress Expressions
Marine vs. Rational BuoyancyMarine Buoyancy Rational Buoyancy
Fbuoy
Fbuoy
conservative estimate of local hydrostatic stresses ( = 0.5, fixed end restraints)
more accurate, direct inclusion of local hydrostatic stresses
Conclusions – Hydrostatic Pressure
API RP 2A-WSD 21st edition and ISO 19902:2007 1st edition have identical code check equations when members are in axial compression with hydrostatic pressure.
For members in axial tension with hydrostatic pressure, the code check equations are different. API RP 2A-WSD 21st edition code check equations provide a reasonably conservative fit to existing test data.
It is recommended to adopt API RP 2A-WSD 21st edition equations for hydrostatic code checks.
Study on 2) Conical TransitionAKADEME Case Study Platform 3
Helideck EL (+) 71.1 ft
Main Deck EL (+) 50.8 ft
EL (+) 22.3 ft
Water Surface EL (+) 10 ft
EL (+) 0 ft
Mudline EL (-) 35 ft
WT=2 in
WT=2 in
• L-3 structure• Free standing caisson• Pile penetration = 110 ft• Pile diameter = 72 inches• Water depth = 45 ft• Clays overlying dense sands
Comparison of Conical Transition Code Check Results
Note: Governing UC ratios are highlighted in red.
• API RP 2A-WSD computes UCs based on tensile strength instead of yield strength for the total stress check.
• ISO 19902:2007 1st edition produces the most conservative result.
• API RP 2A-WSD 21st edition and NORSOK standard produce comparable results.
Maersk FEM Results Compared to Codes- Axial Compression
Maersk report: “Plastic Hinge Model for Tubulars and Conical Transitions. Results of detailed FEM analyses condensed into a beam Model” Rambøll Ref. 862601/340_0001(1)
Conclusions – Conical Transition The UC value for the conical transition member of Platform
3 based on ISO 19902:2007 1st edition is much higher than the UC value based on API RP 2A-WSD 21st edition (nearly twice as high), which indicates that ISO 19902:2007 1st
edition code check equations are more conservative.
API RP 2A-WSD computes UCs based on tensile strength instead of yield strength for the total stress check. Thus, it is recommended to consider adopting the NORSOK conical transition code check equations for the future revisions of API RP 2A-LRFD and ISO 19902.
API RP 2A-WSD 21st edition and API RP 2A-LRFD 1st edition have identical code check equations for conical transitions.
Study on 3) Grouted ConnectionDemand/Capacity API RP 2A-WSD ISO 19902:2007
Transfer Stress only Axial Force
Axial Force&
Torsion
Allo
wab
le T
rans
fer
Stre
ss
Ope
rati
ng C
ondi
tion
Extr
eme
Load
Co
ndit
ion
(including 1/3 allowable stress increase)
without shear keys
with shear keys fg is the lesser of fg,sliding and fg,shear
kred is reduction factor for effects of movement during grout setting;Assume no movement kred = 1
without shear keysℎ� = 0
without shear keys
with shear keys
API RP 2A-WSD vs. 2A-LRFD 1st (w/ Shear Keys)2A-WSD 2A-LRFD 1st
Operating condition
Extreme loading condition
x 1.8
API RP 2A-WSD vs. 2A-LRFD 1st (w/o Shear Keys)2A-WSD 2A-LRFD 1st
Operating condition
Extreme loading condition
0
0
0x 1.8
Case Study 1 – Platform X• Omnidirectional metocean conditions were
applied. operating condition extreme loading condition
0 degree
Ds = 93 ints = 2.5 inDp = 84 intp = 2.5 inL = 44 ft
Assume:h = 0.5 ins = 20 infcu = 5000 psi
Case Study 2 – AKADEME Platform 5• Jacket legs and through piles were assumed to be
grouted in this study.• Omnidirectional metocean conditions were
applied. operating condition extreme loading condition
0 degree
Platform North
Ds = 58.5 ints = 1.25 inDp = 54 intp = 1.75 in
Assume:h = 0.5 ins = 20 infcu = 5000 psiL = 100 ft
Grouted Connection Unity ChecksAPI RP 2A-LRFD 1st vs. API RP 2A-WSD 21st
Base Equation:
API WSD API LRFD 1st
Operating
Condition� + � + � ≤ �� �
1.3� + 1.5� + 1.215� ≤ (0.9 × 1.8)�� �
� . � �� + � . � �� + � . � �� ≤ �� �(normalized)
ExtremeCondition
� + � + � ≤ 1.33�� �
� . � �� + � . � �� + � . � �� ≤ �� �(normalized)
1.1� + 1.1� + 1.35� ≤ (0.9 × 1.8)�� �
� . � �� + � . � �� + � . � �� ≤ �� �(normalized)
Grouted Connection Unity ChecksAPI RP 2A-LRFD 1st vs. ISO 19902:2007 1st
Conclusions – Grouted Connection The grouted connection code check equations in API RP 2A-LRFD
1st edition yield consistent results with those from API RP 2A-WSD 21st edition and ISO 19902:2007 1st edition.
The API RP 2A-LRFD 1st edition has similar formation of interface transfer strength but 1.8 times higher than the allowable transfer stress in API RP 2A-WSD 21st edition.
It is recommended to adopt the API RP 2A-LRFD 1st edition grouted connection code check equations for API RP 2A-LRFD 2nd edition.
Conclusions for ISO/API Jacket CodeWork performed to date supports the approach of Modified
Adoption of ISO 19902 for API RP 2A-LRFD 2nd Edition.Overall, member and joint UCs from ISO 19902 and API
RP 2A-WSD are consistent.Key technical issues studied:
–Hydrostatic pressure–Conical transition–Pile sleeve grout equation
ISO 19902 is currently being updated. API/ISO committees will work closely to harmonize the next editions of API RP 2A-LRFD and ISO 19902.
Presentation Part 2API RP2A Ductility Requirement and Seismic Design Process
Ductile Framing
(source: API RP2A)
Non-Ductile Framing
(source: API RP2A)
API Ductile Design Features - Jackets
Legs and enclosed piles remain elastic for 2 x strength level
Vertical diagonal bracing configured so tension and compression braces get equal shear
No K bracing allowed Braces remain elastic for 2 x strength level in lieu of
above 2 requirements Horizontals between legs at all framing levels with
capacity for load redistribution Vertical diagonals: kL/r<80 D/t<1900/Fy Non-tubulars in vertical frames are AISC
compact sections or meet 2 x strength level
Vertical Bracing Schemes
K Diamond X Diagonal
Comparative Ductility of Vertical Schemes
Effect of Strong Horizontal Planes
API/ISO Earthquake Code Recommendations
API 2EQ is a modified adoption of ISO 19901-2:2004
Two level approach. Extreme Level Earthquake (ELE) – strength design.
Abnormal Level Earthquake (ALE) – ductility check.
Seismic maps for USA offshore in API 2EQ.
Seismic maps for offshore locations worldwide in ISO 19901-2:2004.
Extreme Level Earthquake (ELE)
Platform performs in an elastic manner.
Response spectrum analysis (or time-history).
Code-based or site specific spectra as input.
300 to 500 year Return Period typical.
70% allowable stress increase for API code check.
Do members/joints fail?
Site-specific response spectra based on PSHA and site response analysis may be used instead of the code-based spectra.
Abnormal Level Earthquake (ALE)
Platform typically performs in an inelastic manner.
Time-history analysis (or pushover) to check the “ductility” of the platform.
Site specific ground motion records as input.
3000 year Return Period (or more) typical.
Does the platform collapse?
Global X Component
-1-0.8-0.6-0.4-0.2
00.20.40.60.8
1
0 5 10 15 20 25
Time (sec)
Acc
eler
atio
n (g
)
Global Y Component
-1-0.8-0.6-0.4-0.2
00.20.40.60.8
1
0 5 10 15 20 25
Time (sec)
Acc
eler
atio
n (g
)
Global Z Component
-0.60
-0.40
-0.20
0.00
0.20
0.40
0.60
0 5 10 15 20 25
Time (sec)
Acc
eler
atio
n (g
)
Overall API/ISO Process
Determine Seismic Risk Category (SRC) - higher risk (e.g., manned platforms) = safer design
Determine need for simplified or detailed procedure
Establish Return Period for ALE Establish Return Period for ELE based on the
ductility expected of the platform design Design the platform for ELE (strength design)
and check its ductility/performance for ALE
API/ISO Processfor Ground Motion
(source: ISO 19901-2:2004)
Site Seismic Zones
Seismic Zone Acceleration (g) - 1 sec0 <0.031 0.03-0.102 0.11-0.253 0.26-0.454 >0.45
Example Seismic Map
ISO Map for Central America
(source: ISO 19901-2:2004)
Seismic Risk Category Based on exposure level and seismic zone at the platform site
Note that L2 was removed in API 2EQ (can’t “evacuate” personnel for earthquakes)
(source: ISO 19901-2:2004)
Seismic Design Requirements
(source: ISO 19901-2:2004)
Target Probability of Failure
•Use of Return Periods as a basis for design worldwide can lead to inconsistencies•A 1,000 year RP earthquake in California is different than a 1,000 year RP earthquake in SE Asia•API/ISO uses an alternative approach of a consistent target annual probability of failure, Pf
(source: ISO 19901-2:2004)
Standard API/ISO Spectrum(1,000-yr Return Period)
This shows a “generic” earthquake response spectrum taken from API 2EQ (Nov. 2014). Notice how the “acceleration” that a platform will experience depends on the platform’s dynamic characteristics (periods).
(source: API 2EQ, November 2014)
Scale Factor for ALE Spectrum(Simplified Seismic Action Procedure)
(source: ISO 19901-2:2004)
Correction Factor for ALE Spectrum(Detailed Seismic Action Procedure)
(source: ISO 19901-2:2004)
Seismic Reserve Capacity Factor (Cr) for ELE Spectrum
rC = Variable up to 2.80as demonstrated by analysis.
The recommendations for ductile design in 5.3.6.4.3 are followed and a non-linear static pushover analysis according to API RP 2EQ is performed to verify the global performance of the structure under ALE conditions.
rC = Variable up to 2.00as demonstrated by analysis.
The recommendations for ductile design in 5.3.6.4.3 are followed, but a non-linear static pushoveranalysis to verify ALE performance is not performed.
rC = 1.40
The structure has a minimum of three legs and a bracing pattern consisting of leg-to-leg diagonals with horizontals, or X-braces without horizontals. The slenderness ratio (KL/r) of diagonal bracing in vertical frames is limited to no more than 80 and FyD/Et ≤ 0.069. For X-bracing in vertical frames the same restrictions apply, where the length L to be used depends on the loading pattern of the X-bracing.A non-linear analysis to verify the ductility level performance is not performed.
rC = 1.10 If none of the above characterizations apply.
Minimum ELE Return PeriodsConsequence Category Minimum ELE Return Period
Low Consequence 50Medium Consequence 100
High Consequence 200
•Cr is a function of the expected ductility of the platform design•The intent is to determine a RP that can be used to design the platform elastically, such that it will meet the ALE performance criteria•The Cr factor of 2.0 or higher should be demonstrated by detailed analysis
Summary
Latest API RP2A-LRFD, and ISO 19902 developments were discussed
Some of the succinct features in API/ISO seismic design provisions were also discussed
Technologies in API RP-2A and/or ISO 19902 for designing/building a fixed platform are quite mature
Design code expertise can be leveraged through industry professionals who participated in API/ISO committees
End of PresentationQ & A