Verification of Offshore Wind Modeling Tools through IEA Wind Tasks 23 and 30 ICOWEOE October 31, 2011 Beijing China Amy Robertson, Ph.D. Senior Engineer, NREL NREL is a national laboratory of the U.S. Department of Energy, Office of Energy Efficiency and Renewable Energy, operated by the Alliance for Sustainable Energy, LLC.

Outline What are IEA Tasks Overview of IEA Wind Tasks 23 and 30 Status and results of these tasks

IEA Wind Tasks IEA = International Energy Agency IEA Wind Tasks are cooperative research on issues affecting wind energy The technical results of Tasks are shared among participants. Final reports are often made public to benefit the entire wind energy community

IEA Wind Tasks Country Commitments To be a participant of an IEA Wind Task, your country must first have general membership in IEA Wind. Then, interested countries must join an individual task. This requires a fee to support the operating agents of that task. Spar Concept by SWAY

IEA Wind Tasks Tasks 23 and 30 IEA Wind Task 23 : Subtask 2 of this task was the Offshore Code Comparison Collaboration (OC3) Ran from 2005 to 2009 IEA Wind Task 30: The Offshore Code Comparison Collaboration Continuation (OC4) Continues work of OC3 project Will run from 2010 to 2012 Focus is on offshore wind turbine (OWT) code verification & benchmarking, with emphasis on the support structure Spar Concept by SWAY

The OC3 & OC4 Projects Background Offshore Wind Turbines (OWTs) are designed using aero-hydro-servo-elastic codes The codes must be verified to assess their accuracy OC3 and OC4 approach verification through code-to-code comparisons

The OC3 & OC4 Projects Background IEA Task Wind 30 (OC4) is a follow-on task to IEA Wind Task 23 (OC3) OC4 will continue the examination of more OWT configurations Floating Spar Buoy Monopile Tripod

OC3/OC4 Activities & Objectives Discuss modeling strategies Develop suite of benchmark models & simulations Run simulations & process results Compare & discuss results Assess simulation accuracy & reliability Train new analysts how to run codes correctly Investigate capabilities of implemented theories Refine analysis methods Identify further R&D needs Activities Objectives

OC3 Participants & Codes 3Dfloat ADAMS-AeroDyn-HydroDyn ADAMS-AeroDyn-WaveLoads ADCoS-Offshore ADCoS-Offshore-ASAS ANSYS-WaveLoads BHawC Bladed Bladed Multibody DeepC FAST-AeroDyn-HydroDyn FAST-AeroDyn-NASTRAN FLEX5 FLEX5-Poseidon HAWC HAWC2 SESAM SIMPACK-AeroDyn Simo

Aero-Hydro-Servo-Elastic Capabilities

OC3/OC4 Approach & Phases All inputs are predefined: NREL 5-MW wind turbine, including control system Variety of support structures Wind & wave datasets A stepwise procedure is applied: Load cases selected to test different model features OC3 ran from 2005 to 2009: Phase I: Monopile + Rigid Foundation Phase II: Monopile + Flexible Found’tn Phase III: Tripod Phase IV: Floating Spar Buoy OC4 will run from 2010 to 2012: Phase I: Jacket Phase II: Floating semisubmersible Approach Phases

NREL 5-MW Baseline Wind Turbine Specifications developed for a representative multi-megawatt wind turbine Heavily influenced by the REpower 5M prototype & DOWEC project turbines Properties developed: blade structural & aerodynamic properties nacelle & hub drivetrain tower control system: Also used as UpWind reference model REpower 5M Wind Turbine Gross Properties Chosen for the Baseline Turbine

Load Cases 1.X – Full-System Eigenanalysis 2.X – Rigid Full-system flexibility Elastic response only Compared natural frequencies & damping ratios 2.X – Rigid Rigid turbine Aerodynamics without hydro: Steady & turbulent winds Hydrodynamics without aero: Regular & irregular waves 3.X – Onshore Wind Turbine Flexible tower, drivetrain, & rotor Rigid substructure Aero-servo-elastics without hydro: 4.X – Inverted Pendulum Flexible support structure Rigid tower top Hydro-elastics without aero: Regular & irregular waves 5.X – Full-System Dynamics Full-system flexibility Full aero-hydro-servo-elastics: Steady winds with regular waves Turbulent winds with irregular waves

Output Parameters

Overview of OC3 Project Phases & Status Phase I – Monopile + Rigid Foundation, 20 m water depth Winter 2005 – Summer 2006 Paper presented at Science of Making Torque from Wind, 2007 Phase II – Monopile + Flexible Foundation, 20 m water depth Summer 2006 – Fall 2007 Paper presented at European Offshore Wind, 2007 Phase III – Tripod, 45 m water depth Winter 2007 – Fall 2008 Paper presented at AIAA, 2009 Phase IV – Floating Spar, 320 m depth Summer 2008 – Winter 2009 Paper presented at EWEC, 2010 Final Report – Spring 2010 – NREL Report #TP-5000-48191 Summary journal article under review for Wind Energy Monopile Tripod Spar

OC3 Phase I Results Monopile Modal-based codes predict different 2nd (& higher) eigenmodes than the higher fidelity multibody- & FEM- based codes Codes using full-field wind in polar coordinates predict smoother aerodynamic loads than codes using rectangular coordinates: Because of the method used to generate the wind datasets Differences in implementation of aerodynamic models attributed to variations among the codes in mean turbine loads Differing model discretizations lead to differing code predictions: Most apparent in substructure loads that depend highly on the discretization of hydrodynamic loads near the free surface User error happens

Review of Phase I (Monopile) Results Full-System Eigenanalysis

Review of Phase I Results Onshore Turbine with Turbulent Winds Flexible tower, drivetrain, & rotor Aerodynamics with control system Full-field turbulent wind with: Vhub = 11.4m/s, TI1 = 17.4%

Tripod (Image: J. Nichols, GH) OC3 Phase III Results Tripod Jump in complexity from monopile to tripod: Multiple members Statically indeterminate (loads influenced by relative deflection of members) Nonaxisymmetric Key findings: A fine discretization of hydrodynamic loads is required near the free surface Overlapping regions where structural members join at nodes have a large effect Despite having thin members, shear deflection through Timoshenko beam theory has a large effect on the tripod load distribution Tripod (Image: J. Nichols, GH) Spar Concept by SIWAY

OC3 Phase III Results Eigenanalysis

OC3 Phase IV Results Spar Buoy Jump in complexity from fixed to floating: Low-frequency modes (influence on aerodynamic damping & stability) Large platform motions (coupling with turbine) Complicated shapes (radiation & diffraction) Moorings (new component) Key findings (may only apply to this spar): Radiation damping is negligible; so, codes that apply Morison’s equation are adequate Quasi-static mooring models provide adequate reactions for global response analysis; dynamic mooring models, however, result in more line excitation at higher frequencies Turbine structural flexibilities had an effect on turbine loads, but little effect on spar motions Coupled dynamics issues not fully resolved OC3-Hywind Spar Buoy Spar Concept by SIWAY

Overview of OC4 Status of Phases Phase I: Jacket Nearing completion of analysis Technical paper to be presented at International Offshore and Polar Engineering Conference (ISOPE) in Rhodes, Greece, June17- 23, 2012 Phase II: Semi-submersible DeepCwind semi design will be used Design specification document will be presented at next face-to-face meeting on December 2, 2011

UpWind WP4 Reference Jacket (Images: F. Vorpahl, OC4 Phase I Jacket Title: Verification of simulation codes for a jacket-supported fixed-bottom WT Coordinator: Fraunhofer-IWES Rambøll has kindly agreed to make the UpWind WP4 reference jacket available to OC4 participants: Jacket supports the NREL 5-MW turbine 4 legs, 4 levels of X braces, & mud braces Concrete Transition Piece (TP) & 4 central piles 50-m water depth UpWind WP4 Reference Jacket (Images: F. Vorpahl, Fraunhofer-IWES) Spar Concept by SWAY

Output Locations on Jacket for Response Comparison Transition piece and grout connection to tower

UpWind WP4 Reference Jacket (Images: F. Vorpahl, OC4 Phase I Status Nearing completion of analysis Technical paper to be presented at International Offshore and Polar Engineering Conference (ISOPE) in Rhodes, Greece, June17- 23, 2012 UpWind WP4 Reference Jacket (Images: F. Vorpahl, Fraunhofer-IWES) Spar Concept by SWAY

OC4 Phase II Plans Semi-submersible Title: Verification of simulation codes for a WT on a floating semi-submersible Coordinator: NREL OC4 participants have chosen to use the DeepCwind semi-submersible design A generic, publically available design Wave-tank tested at 1/50th scale in May 2011 Code-to-code comparison results will be published in a conference paper in 2012/2013 Spar Concept by SWAY DeepCwind Semi-submersible

OC4 Coordination & Meetings E-mail coordination Meetings Net-meetings held every 1-2 months Physical meetings held 1-2 times per year 3rd physical meeting held at ISOPE Conference in Hawaii, USA, June 2011 Next meeting at EWEA Offshore 2011 in Amsterdam, Netherlands, November IEA Wind Task 23 and 30 SharePoint website: http://oc4.collaborationhost.net Hosted by NREL Includes meeting presentations, minutes, model and load case descriptions, and simulation results Password required to access site, and certain directories open only to IEA Wind Task 30 members

OC4 Code Validation Experts Meeting OC4 focuses on code-to-code verification; code-to-experiment validation also needed OC4 participants don’t comprise all experts needed to develop field validation plan Separate meeting proposed for 2011 NREL (Walt Musial) will organize the meeting Meeting objectives: Catalogue existing datasets for selected configurations available for code validation Establish methods for collecting new data Make recommendations for R&D agencies that may want to support the effort as a follow-on task Stand-alone report will be published Image: J. ‘t Hooft, SenterNovem

Summary OC3/OC4 aims to verify OWT dynamics codes Benchmark models & simulations established Simulations test a variety of OWT types & model features Code-to-code comparisons have agreed well Differences caused by variations in: Model fidelity Aero-, hydro-, & structural-dynamic theories Model discretization Numerical problems User error Many code errors have been resolved Engineers equipped with modeling experience Verification is critical to advance offshore wind Spar Concept by SWAY Semi-submersible Concept

Thank You for Your Attention Amy Robertson +1 (303) 384 – 7157 Amy.robertson@nrel.gov NREL is a national laboratory of the U.S. Department of Energy, Office of Energy Efficiency and Renewable Energy, operated by the Alliance for Sustainable Energy, LLC.

OC3 Phase II Results Foundation Modeling Foundation designed by applying realistic soil properties & typical design procedures Design made to have a noticeable effect on the system’s dynamic response Needed for facilitation of model testing Nonlinear + depth-dependent p-y model Apparent Fixity (AF) Model Coupled Springs (CS) Model Distributed Springs (DS) Model Soil Profile The p-y model not necessarily valid for transient analysis Simplified models derived to give equivalent response under particular conditions Simplified Models of a Monopile with Flexible Foundation

OC3 Phases II Results Overview All of the results of Phase I also apply to results of Phase II The simplified foundation models can be implemented so as to ensure that the overall response of the system above the mudline is the same under a given set of loading conditions: At least for the lowest system eignemodes Model discretization problems result in higher excitation in the 2nd eigenmodes of the support structure: This is only so when the turbine is not operating, because of the aerodynamic damping while operating Differences in aerodynamic models have more effect on the mean values of loads than on power spectra

IEA Task 23 Organizational Structure

OC4 Project Schedule Spar Concept by SWAY

OC4 Organizational Structure

IEA Wind Task 23 OC3: Phase IV Results Regarding Floating Wind Turbine Modeling NREL – Jason Jonkman MARINTEK – Ivar Fylling Risø-DTU – Torben Larsen GH – James Nichols Anders Hansen LUH – Martin Kohlmeier IFE – Tor Anders Nygaard Acciona – Javier Pascual Vergara UMB – Karl Jacob Maus Daniel Merino NTNU – Madjid Karimirad POSTECH – Wei Shi Zhen Gao Hyunchul Park Torgeir Moan Operated for the U.S. Department of Energy Office of Energy Efficiency and Renewable Energy by the Alliance for Sustainable Energy, LLC

Floating Challenges & Phase IV Model Low frequency modes: Influence aerodynamic damping & stability Large platform motions: Coupling with turbine Complicated shape: Radiation & diffraction Moorings Statoil supplied data for 5-MW Hywind conceptual design OC3 adapted spar to support the NREL 5-MW turbine: Rotor-nacelle assembly unchanged Tower & control system modified Challenges OC3-Hywind OC3-Hywind Model

Phase IV Load Cases

Output Parameters & Results Legend Drivetrain & Generator Loads & Operation 7 Outputs Rotor Blade Loads & Deflections 13 Outputs Tower Loads & Deflections 15 Outputs Environment Wind & Waves 4 Outputs Mooring System Fairlead & Anchor Tensions & Angles 12 Outputs Platform Displacements 6 Outputs Output Parameters (57 Total) Results Legend

Unresolved Issues of OC3 Phase IV Close agreement was not achieved by all codes: What was the reason? The ”effective RAO” load case was somewhat ”academic”: What response charateristic is more relevant? Alternative suggested by IF — RAOs could be derived from irregular time series & cross spectra between excitation & response The stochastic response statistics & spectra are sensitive to simulation length: What length would be more appropriate? How can we eliminate start-up transients from the comparisons?

Limitations of OC3 Phase IV OC3-Hywind platform was considered as a rigid body; no hydro-elastic effects OC3-Hywind platform is simple in shape; only a single member Hydrodynamic radiation & diffraction was negligible in the OC3-Hywind spar buoy Sea current was never considered Few sea states were tested; larger waves may be interesting The relative importance of 2nd versus 1st order hydrodynamics was never assessed The relative importance of dynamic versus quasi-static mooring models was never assessed The influence of platform motion on rotor aerodynamics was never looked at in detail