WIXLIB

API 4F Design Study: ASD‐89 to LRFD‐05Mike Effenberger, P.ESathish Ramamoorthy, Ph.D., P.E.June, 2012; PN 1101213Document No. SES1101213‐PPT‐001,Rev.0aPreliminary subject to QA check

Contents API 4F WorkBackgroundObjectivesDesign Study‐MethodologyModel DetailsAISC‐05 Analysis SpecificationsRig 1 Design Study Results Detailed Design Results Rig 3 Design Study Results Detailed Design Results2

API 4F 3API Funded Study – Check Effect of Load Factors for Drilling StructuresPaper – Code Conversion Issues Going From ASD to LRFD – Draft Provided– Comments Due June 24First Study Meeting – 2‐22‐2012 API 4F Work Group– Mark Trevithick (T&T Engineering) – Chair– Anthony Mannering (Precision Drilling)– Marcus McCoo (NOV)– Paul Landis (Lee C Moore) First Load Factors – ASCE7 Models Received (last): 03/05/2012 Finish Work Date: 11/05/2012

API 4F Design Study‐Status6‐13‐2012 Paper – Issued as Draft (Close 6-24-2012) Issue to API End of JuneRig 1 Boot-Strap Mast – Complete (Pending Review) ASD-89 (StruCAD & SAP) ASD-05 – 1st Order & ASD-05- P-Delta LRFD-05Rig 3 Derrick ASD-89 & LRFD-05Rig 2 Mast & Substructure In progress – Duplicate Plate Element Issues (Removed)Rig 4 Dual Derrick – Duplicate Element Issue (Merging Elements)Rig 5 Workover – Staad Model – Transfer Issues (Working withDesigner) 4

Background API 4F, Specification for Drilling and Well Servicing Structures API 4F committee is interested in going from the AllowableStress Design to the latest strength design Allowable StrengthDesign (ASD)/Load and Resistance Factor Design (LRFD)provision in AISC 360‐05 specification. Latest AISC specification is AISC 360‐10. The 2010 editionsupersedes and is an update of the 2005 edition. The 2010edition is yet to be supported by Computer Programs.5 states requirements and gives recommendations for suitable steelstructures for drilling and well‐servicing operations API 4F specifies the steel structures to be designed in accordance withAllowable Stress Design/Elastic design per AISC 335‐89

Background In the LRFD design methodology, design results areaffected by the load factors selected for the individualloads (self weight, hook load, and environmental loads) For building design, load factors and load combinations arespecified in ASCE 7, Minimum Design Loads for Buildings andOther Structures The load factors are generally estimated using thestatistical data for the individual loads and reliabilityconcepts Statistical data for Hook load is not readily available. Therefore,the load factors given in ASCE 7‐05 will be used for the designstudy (first case evaluation)6

Objectives Paper documenting the experience of other codes/specification while changing from ASD to LRFDmethodologyPerform design study on six structures (workover,mast, derrick & substructure) designed using the ASDand LRFD design provisions.Hook load will be considered as live load with a loadfactor of 1.6 (ASCE 7, First Case Evaluation) Decided by the API 4F work group on 02/22/20127

Methodology Design performed using the Allowable Stress 8Design/Elastic Design provisions in AISC 335‐89 willbe the base case.The structures will be evaluated according to AISC335‐89 ASD and LRFD design provisions per AISC360‐05 specificationThe design results between the Elastic design andStrength Design will be compared for evaluation

Derrick and Mast Models Provided toSES for Design StudyElevationNo of Setfrom WaterLines Back HeightLineOperating Unexpected Expected(kip) (ft)(ft)Reference Wind Speed(knots)ModelNameTypeHookLoad(kip)Rig 1Mast44110xxx92427093102.5Rig 2Mast &Substructure750125001603271.795.6Land RigRig 3Single Derrick150014xxx1955010011570Dual Derrick25001500(Aux)24265.1xxx10083.7104256075Rig 4Rig 5** Workover Mast2501614 1750(Aux)6xxxNotes:Analytical models for Rig1 to Rig 4 were developed in StruCADFor Rig 5, Analytical model was developed in STAAD9Land Rig

Rig 1441 kip Hook Load Mast10

Rig 2750 kip Hook Load Mast & SubstructureZX11Y

Rig 31500 kip Hook Load Single DerrickZX12Y

Rig 42500 kip Hook Load Dual Derrick13

Rig 5250 kip Hook Load Workover Mast14

Software Details SAP2000 Ultimate 15.1.0 from Computers and Structures, Inc (CSI) was used for the design study.SAP is a integrated software for structural analysisand design P‐Δ with large displacements Buckling analysis Design �10 (Later in 2012)

StruCAD to SAP2000 Conversion StruCAD models were imported in SAP. The imported 16models were compared with the StruCAD models.In general, most of the model features are importedwithout any issues. However, there are few StruCADcommands/features that are not imported in SAP.These are manually edited in the imported model.SAP technical support and development team werenotified about the import issues.

StruCAD to SAP 2000 ConversionOutstanding issues Effective Length Factors not imported Effective length specified/used in the StruCAD model isentered as “Design Overwrites” in SAP model Wind Loads/Area Cards are not imported Wind loads are converted to Member and Joint Loads inStruCAD and imported in SAP Acceleration Loads Discrepancy in the base reactions for the accelerationloads in the StruCAD and SAP models. Therefore, theacceleration loads are converted to Member/joint loads inStruCAD and then imported in SAP17

StruCAD to SAP 2000 ConversionDesign issues Cm factor Difference in calculating the Cm factor between StruCADand SAP Brace Spacing StruCAD has option to input to the brace spacing forestimating the effective length. SAP does not have anequivalent parameter to modify18

Design Basis AISC‐1989 Specification Allowable Stress Design (ASD‐89) AISC‐2005 Specification Load and Resistance Factor Design (LRFD‐05)19

Design BasisAllowable Stress Design‐89 The imported models were run in SAP and comparedwith the StruCAD results In general, the member forces are comparable to theStruCAD member forces. For some members, there is asmall discrepancy in the minor and major axis moments. Allowable Stress Design (ASD‐89)– Static linear analysis– With 1/3rd increase in allowable stress for expected andunexpected load case20

Design Basis Allowable Strength Design‐05For Rig 1 OnlyAllowable Strength Design (ASD‐05)––––Effective length methodSecond order analysis method (includes p‐Δ and p‐δ)Without 1/3rd increase in allowable for wind load combinationsAPI 4F load combinationsCase1a1b23a3b*4521Design edUnexpectedErectionTransportationDead Load(%)100100100100100100100Hook Load Rotary 10001000Setback Load(%)1001000100‐0100Environmental Loads(%)100100100100‐100100Note: Load combination 3b is for Earthquake loads and was not includedin this study

Design BasisLoad and Resistance Factor Design ‐05 Load and Resistance Factor Design (LRFD‐05)––––Effective length methodGeneral second order analysis method (includes p‐Δ and p‐δ)Without 1/3rd increase in allowable for wind load combinationsHook load is considered as live load with 1.6 load factor similar toASCE 7‐05 live loadsCase1a1b23a3b45Design Loading Dead Load Hook Load Rotary Load Setback Load Environmental 0100Note: Other loads should be classified (for ex loads in Rig 3 Model22

Comparison of Analysis MethodsAISC‐05 SpecificationDirect AnalysisMethodEffective LengthMethodFirst‐Order AnalysisMethodSpecification ReferenceAppendix 7Section C.2.2aSection C.2.2bLimits on irst‐OrderReduced EI and EANominal EI and EANominal EI and EANotional lateral loadYesYesYesColumn effective lengthK 1Sidesway bucklinganalysisK 1Type of analysisMember stiffnessNote: In AISC‐10 specification, Direct Analysis is moved to Section C.Effective length and Firs‐Order Analysis Methods are moved to Appendix 723

Comparison of Analysis MethodsDirect AnalysisMethodEffective LengthMethodFirst‐Order AnalysisMethodSpecification ReferenceAppendix 7Section C.2.2aSection C.2.2bLimits on irst‐OrderReduced EI and EANominal EI and EANominal EI and EANotional lateral loadYesYesYesColumn effective lengthK 1Sidesway bucklinganalysisK 1Type of analysisMember stiffnessNot considered in this study24

Effective Length Method DetailsLinear Load combinationNo P‐DeltaAmplified First OrderK 1Force effects are amplified by B1and B2 factor to account for P‐Delta**Effective LengthMethodNon‐linear load combinationP‐deltaGeneral SecondOrderEffective length from StruCADmodel (side sway bucklinganalysis was not performed)Notional lateral load for gravityload combinations is notincluded** B1 and B2 factors account for the P‐δ and P‐Δ. For the current API 4Fstudy, to study the effect of changes in the specification for memberstrength, the B1 and B2 factor of 1.0 was used. Comparison of design resultsbetween the ASD‐05 (First order) and ASD‐89 will show the effect of changesin specifications.25

Rig 1 Design Study Results(Boot Strap Mast) Total number of Members 347 (excluding 16 dummy members) SAP classifies the members Beams (all elements parallel to the X‐Y plane)– 184 Brace (all elements not classified as beams or columns)– 85 Column (all elements parallel to the Z‐axis)– 7826

Rig 1441 kip Hook Load MastRig 127Rig 1Beams184Rig 1Braces85Rig 1Columns78

Rig 1‐ Design Study Results441 kip Hook Load MastMemberswith unityratio 1.0Type tress Design(ASD‐89)StruCADAllowableStress Design(ASD‐89)SAPAllowableStress 5)AmplifiedFirst OrderMethod**AllowableStrength Design(ASD‐05)Load andResistanceFactor Design(LRFD‐05)(ASCE erFirst‐OrderSecond‐Order(P‐Δ and P‐δ)Second‐Order(P‐Δ and P‐δ)YesYesNoNoNoNoNote: **The amplified first order method is modified (B1, B2 1) tocompare the difference in design results between ASD ‐89 and ASD‐05specifications

Rig 1 Design Study ResultsUnity Check StatisticsSAP Models‐ ASD‐89 and LRFD‐05No of members with UC 1Maximum UCAverage UCBeamBraceASD‐89 LRFD‐05 % change ASD‐89 LRFD‐05 % change12171.161.05‐9.14%1.092.28 109.15%0.490.549.16%0.390.4412.60%ColumnASD‐89 LRFD‐05 % change12171.191.3614.53%0.720.753.69%Notes1.SAP ASD‐89 results are with 1/3rd increase in Allowable Stress for Expected and UnexpectedStorm2.LRFD‐05 results are without 1/3rd increase. Load factor for Hook Load is 1.63.Even without the 1/3rd stress increase for LRFD‐05, there is no significant increase in number ofmembers exceeding the unity check ratio4.The significant change in UC for beam member from ASD‐89 to LRFD‐05 is due to the differencein the interaction equation for doubly symmetric members in LRFD specification5.Detailed results for the highest UC members for beam, brace and column are provided indetailed results section29

Rig 1‐ Design Study ResultsASD‐89 StruCAD and SAPASD‐89‐StruCAD Model13 members with UR lumnBraceBraceBeamBeamBeamBeam30Member No Load Comb Unity .109339181.005340220.851341200.941With 1/3rd Stress Increase forExpected and Unexpected StormASD‐89‐SAP model14 members with UR 1.StruCAD model Importinto SAP2000Design withAISC‐89 specificationwith1/3rd Stress IncreaseMember 79284285337339340341Load SCAD24SCAD24SCAD18SCAD24SCAD18SCAD21SCAD20With 1/3rd Stress Increase forExpected and Unexpected StormUnity 1.0890.9870.4560.482

Rig 1‐ Design Study ResultsASD‐89‐SAP Model14 members with UR mnColumnColumnBraceBraceBeamBeamBeamBeam31Member 278279284285337339340341Load AD20Unity 1.1640.9611.0890.9870.4560.482With 1/3rd Stress Increase forExpected and Unexpected StormASD‐89 SAPAISC‐89 specificationDesign resultswithout the1/3rd Stress IncreaseASD‐89‐SAP model17 members with UR 1.Member 278279284285337339340341Load AD20Unity 1.2311.3061.0890.9870.4560.482Without 1/3rd Stress Increase forExpected and Unexpected Storm