Renewable Industries: From Blade to Grid Ian Jones* … UK/staticassets... · Renewable Industries: From Blade to Grid ... Inc. Proprietary Velocity Magnitude Case Studies: ... –Unsteady

© 2010 ANSYS, Inc. All rights reserved. 1 ANSYS, Inc. Proprietary© 2010 ANSYS, Inc. All rights reserved. 1 ANSYS, Inc. Proprietary

ANSYS Software and the

Renewable Industries:

From Blade to Grid

Ian Jones*

ANSYS UK Technical

Services

*Also visiting Professor, Nottingham University

© 2010 ANSYS, Inc. All rights reserved. 2 ANSYS, Inc. Proprietary

Contents

• Background: Renewable Industries

– Challenges and Opportunities

• Simulation Background

• Simulation Overview: Component Modelling

• Examples

– Wind, onshore and offshore

– Tidal

– Wave

– Hydro

• Concluding Remarks


Challenges in Renewable Energy

• Major opportunities, eg Offshore Wind Energy– This round of development takes Scotland’s total projected

offshore wind capacity to over 11 GW by 2020, drawing in over

£30bn of investment . This could create around 20,000 jobs– Jenny Hogan, Wind Policy Manager Officer, Scottish Renewables,

http://www.scottishrenewables.com/news/scotlands-offshore-wind-power-20000-jobs-30bn-inve/

• Major Barriers

– Increase efficiency, - bigger, better

– Cost Reduction

– Risk Reduction

• Predictability / Bankability /ROI

• Reliability

• Scale Up


Wind, Tide, Wave and Hydro

• ‘Highly Energetic’ sources of energy

• Wind, tide and waves

– Huge Potential

– Innovative

– Immature technology

– Fluctuations in availability and demand

– Significant forces

– Fatigue

• Simulation has a major role in fulfilling the

potential of the technology


In-Depth Technology

Spanning Multiple Domains

Conduction

Convection

Radiation

Phase Change

Mass Transport

More…

Thermal

Compressible

Incompressible

Laminar Flow

Turbulence

Multiphase Flow

Non-Newtonian Fluids

More…

FluidsTech

nic

al D

ep

th

Steady-State, Transient, Harmonic & Modal

Linear & Nonlinear

Technical Breadth

Quasi static (Low Freq)

Full Wave

Joule Heating

Eddy currents

Current flow

Circuit Coupling

More…

ElectromagneticsStructural

Large Displacements

Finite Strain

Contact

Multibody Dynamics

Random Vibration

Implicit & Explicit

More…

Tet/Prism

Hex/Hex Core

Structured

Unstructured

Multi-zone

Body-fitted Cartesian

Patch Independent

More…

Meshing


Design – Engineering – Manufacturing - Operation

Engineering Simulation for

WIND ENERGY

ElectronicFluids OffshoreStructural &

Thermal


Engineering Simulation Applications

Wind Turbine Components

• Blades

• Rotor

• Pitch

• Brake

• Low speed shaft

• Gear box

• Generator

• Controller

• Anemometer

• Wind vane

• Nacelle

• High-speed shaft

• Yaw drive

• Yaw motor

• Tower

• Foundations

Source: DOE Office of Energy Efficiency and Renewable Energy

http://www1.eere.energy.gov/windandhydro/wind_how.html

http://www1.eere.energy.gov/windandhydro/wind_how.html


ANSYS CFD for Wind Energy




Aerodynamic Analyses

Airfoils

Blades

Full rotor assembly

Acoustics


Aerodynamic Blade Design


Case Study: Aerodynamic Performance

AoA = 1°

Pressure (Cp) Distribution S809 Airfoil


Suction Side Pressure Side

Case Study: Aerodynamic Performance

Transition Locations S809 Airfoil


Transition: 3-D NREL Wind

Turbine


Transition: 3-D NREL Wind

Turbine


3D Separation on Wind Turbine


Blade Design

Courtesy of IMPSA

SUCTION PRESSURE


Blade Design

Courtesy of IMPSA

TIP VORTICES


Velocity Magnitude

Case Studies:

Free-spinning horizontal-axis

6-DOF Driven Rotation




Tower Design

• Challenges

– Structural reliability

– Minimize weight and cost

– Wind-rotor induced vibration

• Benefits of simulations

– Effect of tower in efficiency

– Fluid-structure interaction

effects

– Unsteady solution

Photo ©Michael Utech


• Main aspects

– Complete wind

turbine taken on

account

– Rotating geometry

– Steady &

unsteady states

– FSI coupling:

Transfer of forces

on the structure to

ANSYS Structural

– Driven Rotation

Tower Design


Engineering Simulation for

Wind Energy


Nacelle Cooling

• Challenges

– Ensure effective cooling under all

environmental conditions

– Complex geometries & many details

– Many parameters:

o Fan positions & number

o Positions of electrical devices

o Outside temperature &

incoming sun radiation• Benefits of simulations

– Virtual prototyping of different

cooling solutions

– Less trial & error

– Reduce thermal peak loads on

generator, gear, transformer, etc.

– Pre-identify “problem” regionsCourtesy of

GAMESA


• Main aspects

– Reliable turbulence models

– Near wall treatment of boundary

layers

– Coupled simulation of heat

transfer in fluid and solid regions

– Radiation:

o Between surfaces

o Solar Load model

– Electromagnetic thermal sources

can be imported

– Results can be transferred to

structural analysis

Nacelle Cooling


Acoustics

• Challenges

– Aero-acoustic noise

– Coupling of different

noise source

– Community acceptance




Site Selection

• Challenges

– Steep terrain, mountains, forests

– Predict wind behavior

– Varying wind directions and

speeds

• Benefits of simulations

– Power Estimates

– Optimize turbine placement

– Wind speed & turbulence

prediction over complex

terrain


• Main aspects

– Geometry import

– Meshing efficiency

– Turbulence models

– Automatic: WindModeller

Courtesy of Ian Jones, Christiane Montavon, Chris Staples, ANSYS UK

Site Selection


• WindModeller

WindModeller

Work performed by Ian Jones, Christiane Montavon, Chris Staples, ANSYS UK

– Inputs

o Topographic data

o Grid resolution

o Wind conditions

– Outputs:

o Report in html format

o Streamlines

o Plots at constant heights

o Numerical data tables

o Export to Google Earth

Image: © Crown Copyright 2008, License 100048580

Site Selection


Blacklaw Wind Farm

• Central Scotland

– Former open cast coal site

• Operated by Scottish Power Renewables

• Largest operating windfarm in the UK (Jan 2006),

• 54 Siemens Turbines

• Total installed power capacity of 125 megawatts (2.3 MW each)

• Small height variations (170m) across farm

Map Image: Ordnance Survey © Crown Copyright 2008, License number 100048580

C. Montavon, I. Jones, C. Staples, C. Strachan, I. Gutierrez, 2009,

Practical issues in the use of CFD for modelling wind farms,

http://www.ewec2009proceedings.info/allfiles2/70_EWEC2009present

ation.pdf


Multiple Wakes Example:

Wind Farm

Wind speed at hub height, wind direction 210

Without wind turbines With wind turbines


Comparison with Data:

• Wake model significantly improves the prediction of

wind speed distribution on site.

– Wakes persist for several kilometres

0

0.2

0.4

0.6

0.8

1

1.2

1.4

T08 T35 T40 T03 T22 T17 T37 T09 T07 T19 T55

Turbine name

Win

d s

pe

ed

at

hu

b h

eig

ht/

win

d s

pe

ed

at

ma

st

82

data, average

Simulation

5/18/2006 3:30

Without wakes

0

0.2

0.4

0.6

0.8

1

1.2

1.4

T08 T35 T40 T03 T22 T17 T37 T09 T07 T19 T55

Turbine name

Win

d s

pe

ed

at

hu

b h

eig

ht/

win

d s

pe

ed

at

ma

st

82

data, average

Simulation

5/18/2006 3:30

With wakes


Effect of Forestry, Normalised Velocity

With wakes With wakes and forest canopy


Effect of Forestry

• Modelling forest canopy improves prediction for some machines

(green circle) but degrades prediction for some other (red circle)

• Most likely cause: too coarse representation of forest canopy

• Other possible issue: use of single loss coefficient, rather than

using spatially varying leaf area index

0

0.2

0.4

0.6

0.8

1

1.2

1.4

T08 T35 T40 T03 T22 T17 T37 T09 T07 T19 T55

Turbine name

Win

d s

pe

ed

at

hu

b h

eig

ht/

win

d s

pe

ed

at

ma

st

82

data, average

Simulation

5/18/2006 3:30

With wakes

0

0.2

0.4

0.6

0.8

1

1.2

1.4

T08 T35 T40 T03 T22 T17 T37 T09 T07 T19 T55

Turbine name

Win

d s

pe

ed

at

hu

b h

eig

ht/

win

d s

pe

ed

at

ma

st

82

data, average

Simulation, forest canopy

5/18/2006 3:30

With wakes and forestry


Power Prediction

• Further Work

– More detailed forestry and met data

– Annual Capacity Factor vs SCADA data

• Normalised Power Outputs• Acknowledge Scottish Power Renewables and SgurrEnergy

R. Spence, C. Montavon, I. Jones, C. Staples, C.

Strachan, D. Malins, 2010, Wind modelling evaluation

using an operational wind farm site,

http://www.ewec2010proceedings.info/allfiles2/517_E

WEC2010presentation.pdf


An Suidhe

• Resource assessment

• Multiple met masts

• Correlating results

between masts

• Include atmospheric

stability model

• Small changes to flows

• Major changes to

turbulence

An Suidhe Wind Farm CFD vs measurement




• Main aspects

– Optimization: Turbines position

– Hierarchy of Models:

o Detailed geometry

o Virtual Blade Model

o Actuator Disk Model

– Automatic: WindModeller

– Energy production

From: T. Hahm, J. Kröning, In the Wake of a Wind Turbine, Fluent News, Spring 2002, http://www.tuev-nord.de/downloads/tnind_cttfd_lit_wind_2002_wea_wake.pdf

Wind Farm Design


WindModeller• WindModeller

– Wind turbine represented by actuator disk

– Wind turbine orientation parallel to wind direction at inlet

– Automatic mesh refinement in rotor volume

Initial mesh

1st refinement

2nd refinement

Final mesh

Work performed by Ian Jones, Christiane Montavon, Chris Staples, ANSYS UK

Wind Farm Design


Validation: Wake Modelling

• Many Examples

• Eg Comparison

of actuator disk

model with

Galion Lidar,

Whitelee Wind

FarmDinwoodie, Quail, Clive,

Strathclyde Uni and

SgurrEnergy, EWEA 2011


Modelling Large Arrays:

Horns Rev

•8x10 WT

•Diameter of 80m

•Hub height of 70m

•Wind turbine spacing: 7 diameters

•Domain size:

•10 km radius

•1.0 km height

•Wind turbine thrust curve: Vestas

V80 WT

•ABL boundary layer profiles at

inlet

geou

z

zln

uu ),

~(min

0

*

z

u~

3

*

0,max~ zzzz ground

22

3inletrefTIuk

22

*

k

uC


Results at hub height

Horizontal velocity Turbulence intensity

Uref = 10 m/s at 70m, z0 = 0.0001m, upstream TI = 6%

Wind direction: sector 285


0

0.2

0.4

0.6

0.8

1

1.2

1 2 3 4 5 6 7 8 9 10

Norm

alis

ed P

ow

er

Turbine Group

10m/s 270° 2° bin

Model Ec

UpWind

0.2

0.4

0.6

0.8

1

1.2

1 2 3 4 5 6 7 8 9 10

Norm

alis

ed P

ow

er

Turbine Group

10m/s 270° 10° bin

Model Ec

UpWind

0.4

0.6

0.8

1

1.2

1 2 3 4 5 6 7 8 9 10

Norm

alis

ed P

ow

er

Turbine Group

10m/s 270° 30° bin

Model Ec

UpWind

Normalised power down a row

• sector 270

• Reasonably good prediction

– Tendency for over-estimation of array

losses

– Good prediction of slope down the row

• Consistent for various bin sizes


0

0.2

0.4

0.6

0.8

1

1.2

1 2 3 4 5

No

rma

lis

ed

Po

we

r

Column

30 deg bin

Model Ec

Measured Data

Upper 25%

Lower 25%0

0.2

0.4

0.6

0.8

1

1.2

1 2 3 4 5

No

rma

lis

ed

Po

we

r

Column

10 deg bin

Model Ec

Measured Data

Upper 25%

Lower 25%

North Hoyle:

Normalised power down a row

• Very good agreement with power data for both bin sizes

• Absolutely blind test case!

Uref = 10 m/s at 67m, z0 = 0.0001m, upstream TI = 7%

Wind direction: sector 260


Electro-Magnetics

Power Generation Unit

Wind Turbine

Generator

Power Meters

Transformers

Electrical Network

Not Included:

Reactive Power Compensation

Inverter


drive signal forthe converter(voltage)

Q Current Controller

D Current Controller

drive signal forthe converter(voltage)

Actual ID

Ref ID

converter

regulator

regulator

converter

Ref IQ

actual IQ

regulator

Ref Q

Actual Q

regulator

Q Power Controller

P Power Controller

Actual P

Ref P

0

0

0

0

0

0

A1

B1

C1

N1

A2

B2

C2

N2

ROT1

ROT2

w+W

+

WM1

W

+

WM2

W

+

WM3

W

+

WM4

W

+

WM5

W

+

WM6

w

+

ICA:

FML_INIT1

EQU

FML4

STATE_1140

SET: SWA1:=0

SET: SWB1:=0SET: SWC1:=0

STATE_1139

SET: SWA1:=1


STATE_1138

SET: SWA1:=1


STATE_1137

SET: SWA1:=1


STATE_1136

SET: SWA1:=1


STATE_1135

SET: SWA1:=0


STATE_1134

SET: SWA1:=1


STATE_1133

SET: SWA1:=0


STATE_1132

SET: SWA1:=0


STATE_1131

SET: SWA1:=1


STATE_1130

SET: SWA1:=1


STATE_1129

SET: SWA1:=1


STATE_1128

SET: SWA1:=0


STATE_1127

SET: SWA1:=0


STATE_1126

SET: SWA1:=0


STATE_1125

SET: SWA1:=0


STATE_1124

SET: SWA1:=0


STATE_1123

SET: SWA1:=1


STATE_1122

SET: SWA1:=0


STATE_1120

SET: SWA1:=0


STATE_1119

SET: SWA1:=0


STATE_1118

SET: SWA1:=0SET: SWB1:=1

SET: SWC1:=0

STATE_1117


SET: SWC1:=0

STATE_1116


SET: SWC1:=1

STATE_1115


SET: SWC1:=1

STATE_1114


SET: SWC1:=1

STATE_1113


SET: SWC1:=0

STATE_1112


SET: SWC1:=0

STATE_1111


SET: SWC1:=0

STATE_1110


SET: SWC1:=0

STATE_119


SET: SWC1:=0

STATE_114


SET: SWC1:=1

STATE_113


SET: SWC1:=0

STATE_2_2


SET: SWC1:=0

STATE_1121


SET: SWC1:=0

STATE_1_8STATE_1_7STATE_1_6STATE_1_5STATE_1_4STATE_1_3STATE_1_2

STATE_118

STATE_117

STATE_116

STATE_115

STATE_2_1

STATE_1_1

STATE_Flexible1

2L3_GTOS

g_r1

g_r2

g_s1

g_s2

g_t1

g_t2

TWO_LVL_3P_GTO1

C2

B6U

D1 D3 D5

D2 D4 D6

B6U1

+

V

VM1

A

B

C

G(s)

gs4

G(s)

gs3

G(s)

gs2

G(s)

gs1

I

GAIN

sum3

sum2

GAIN

I

sum5

GAIN

I

G(s)

gs5

G(s)

gs6

I

GAIN

sum8

0.00 100.00 200.00 300.00 400.00 500.00 600.00Time [ms]

-400.00

-200.00

-0.00

200.00

307.39

Y1

Curve Info

rotor_current_dTR

rotor_current_qTR

target_rotor_current_DTR

target_rotor_current_QTR

0.00 100.00 200.00 300.00 400.00 500.00 600.00Time [ms]

-50.00

-25.00

0.00

25.00

43.09

Y1

[k]

Curve Info

PTRintgain='2' pgain='0.9'

QTRintgain='2' pgain='0.9'

PRTRintgain='2' pgain='0.9'

QRTRintgain='2' pgain='0.9'

0

0

0

0

EQU

ICA:LIMIT

lmt1

100 %

11

2

T

VSI_3ph_avg

VSI3ph_A1

A_1

B_1

C_1

A_2

B_2

C_2

U3

stator_connect5

A

B

C

+

V

VM1

B6U

D1 D3 D5

D2 D4 D6

B6U1

C2

EQU

ICA:

A1

B1

C1

N1

A2

B2

C2

N2

ROT1

ROT20

0

0

0

EQU

ICA:LIMIT

lmt1

100 %

11

2

T

VSI_3ph_avg

VSI3ph_A1

A_1

B_1

C_1

A_2

B_2

C_2

U3

stator_connect5

A

B

C

+

V

VM1

B6U

D1 D3 D5

D2 D4 D6

B6U1

C2

EQU

ICA:

A1

B1

C1

N1

A2

B2

C2

N2

ROT1

ROT2

Power Electronic:

Complete Turbine System

DQ Power and Current Controllers

Space Vector ModulatorRectifier Inverter

Generator

Electric Grid

Power DeliveredGenerator Current

Wind Source


Electro-Mechanical

Co-simulation with Rigid Body Dynamics

Transient Structural (RBD) Crankshaft ModelANSYS Mech.

Motor Shaft ROM

Control

Inverter/Converter

PM Motor

Throttle ControlAutomotive Drive train

Simulink

Simplorer

ANSYS RBD


R2

R4

Control

Wind Profile

Offshore Windturbine System Modeling

Position/Motion Control

Structural

Mooring

CFD


shielding mooring systems

FPSO & TLP Concept

self-installing platform – Arup Energy

cargo lowered onto vessel offloading operation – SBM

ANSYS AQWA: Linear Wave Analysis

- Typical Applications


ship & landing craft jacket launchlifting operation – AMOG

truss spar – PI-RAUMAfloatover stinger

Offshore Transport and Installation


Diffraction Analysis

• Diffraction/Radiation

multi-body interaction

analysis

• 2nd Order QTF’s (Full &

Newman

approximation)

• Import of external

calculated values /

editing of

hydrodynamic

database

– Added mass

– Damping

– RAOs

– Drift coefficients


Dynamic response in frequency

domain

• Significant motions at low frequency / wave frequency in frequency domain

• Rapid analysis using linearised parameters of mooring systems

• Graphs for response spectra / RAOs & other parameters


ANSYS ASAS

- Coupled Wave-Structures

• Coupled hydro-elastic analysis for

tubular framed structures

– Fully coupled hydrodynamic loading

with non-linear analysis capability

– Automatic computation of

hydrodynamic damping

– Regular and irregular waves

– Ability to take AQWA RAO results

as time history loading

• Coupling to other packages


ANSYS ASAS-Offshore

• ASAS-Offshore

– Spectral and deterministic fatigue

of jacket structures

– Time history fatigue analysis for

transient load conditions

– Seismic loading

– Pile-structure interaction


• CFD Features that may be useful in the Marine Renewables Sector

– Flow Visualisation• Better understanding

• Drag calculation

– Turbine-specific tools• Multiple Frames of Reference

• Performance and power extraction

• Cavitation modelling

– Free Surface models• Simple wave generation

• Wave/body interactions

– Dynamic response• Rigid Body 6dof solutions Eg. Bobbers

• Added mass and damping calculation

– Fluid-Structure Interaction• Fatigue and stress response to fluid load

• Survivability

Marine Renewables: CFD

Courtesy of Cardiff University


Turbine CFD – in good shape

• Wells Turbine– It should be noted that CFD

predictions of Wells turbine performance are comparable to measured data until the turbine stalls, after which they diverge.

– Carbon Trust Report Oscillating Water Column Wave Energy Converter Evaluation Report

• Cycloidal Turbines

• Kaplan / Pelton / Francis

• All with good CFD results

• Similar for other Hydro Power components

– Penstock, spillways, fish traps..

• Support structures, ballasts, local effects can be important


ANSYS Turbo System

• Turbo-specific Pre-Processing

– Multiple component

– Multiple passage

– Automated setup

• Turbo Specific Post-Processing

– Blade-to-Blade plots

– Meridional views

– Performance

– Averaging

– Blade load charts

– Automated Reporting

Courtesy PCA Eng.


Turbine Stresses

• 1-way force calculations for FSI now routine

– Few button presses

– Full 2way FSI coupling to structural solver• Vertical Axis Turbine

• Wells Turbine Calculations by WaveGen, Tease et al


Free Surface Flow

• Free surface flow

– separated multiphase flow

– fluids separated by distinct resolvable

interface / Volume of Fluid Method

– examples: open channel flow, flow around

ship hulls, water jet in air (Pelton wheel),

tank filling, etc. Pelamis


Cavitation

• Explicit cavitation modelling to predict reduction in performance


Case Studies: Marine

• Oscillating Water Columns

• Tidal Turbine

• Added mass and damping calculations for use in

simpler models


OWC principle

• Waves in sea generate

oscillation in vertical duct

• Resonance occurs if duct

diameter and length are

carefully chosen

• Resonance can increase the

wave height significantly

• Cylinder can be on sea-bed,

or at surface

Reference: Lighthill, J., 1979, „Two-dimensional analyses related to wave-energy

extraction by submerged resonant ducts‟, J. Fluid Mech, 91, part 2, 253-317.


OWC principle

• Compression

chamber above

OWC

• Energy can be

harnessed (e.g. via

Wells turbine)

Source:http://news.bbc.co.uk/1/hi/sci/tech/1032148.stm


Vertical cylinder on sea-bed

• Pressure distribution at

resonance

– Amplitude elevation in

cylinder


Wave-piercing design

• Air Pressure variation

inside cylinder


Air movement trough the hole at the

top of the OWC

Wave Piercing Design


Tidal Turbine Example

• Geometry of tower, support wing and pylon created in BladeModeler

• Turbine Blades modified from NREL Wind Turbine

• Tower Height

– 40 meters

• Span Length of Support Wing

– 20 meters

• Rotor Diameter

– 20 meters

• Number of Blades

– 2


Post-Processing

• Pressure

contours on

blade and tower

surfaces

• Streamlines of

velocity

• Dye injection

from tip of

turbine

© 2010 ANSYS, Inc. All rights reserved. 70 ANSYS, Inc. Proprietary© 2010 ANSYS, Inc. All rights reserved. 70 ANSYS, Inc. Proprietary

Added-Mass and Damping Calculation


Added-Mass and Damping

• When simulating floating bodies, or mooring systems,

some 3D-panel method codes and multi-body

dynamics codes require additional coefficients in

order to get an accurate response.

• The effect of these coefficients is implicitly included in

full CFD analyses – sometimes, however, this is not

practical

– Eg. Simulations of motion in irregular waves

• Sometimes coefficients can be estimated for simple

geometry

• For complex geometry we can calculate them quickly

using CFD


Mudmat Example

• Lowering of Mudmat to seabed

– Need added mass and damping for

accurate dynamics simulation

• Perform Transient CFD calculation

– Separate horizontal and vertical motion

prescribed

– Sinusoidal moving mesh

– Simulation duration of 3-5 cycles only


Mudmat Example

• Results analysed in CFD-Post

– Coefficients extracted from amplitude

and phase of reaction force plot

– Also examined:

• Effects of holes in geometry

• Effect of proximity to sea bed

Sway Motion

Heave MotionPeriod

(s)

Amplitude

(cm)

Calculated Value

(no holes)

Model Tests

(4 holes)

Added Mass in Heave

6.0

0.375

76.33

88.0

6.0

0.5

77.17

90.0

6.0

0.75

81.38

92.0

7.0

0.25

76.4

-

Period

(s)

Amplitude

(cm)

Calculated Value

(no holes)

Model Tests

(4 holes)

6.0

0.375

17.91

12.0

6.0

0.5

19.55

13.5

6.0

0.75

18.86

18.0

Damping in Heave


• Minimize development, warranty & liability costs

• Innovative & higher-quality products

• Dramatic time-to-market improvement

Simulation Driven Product Development:

Delivering Simulation Solutions For

Renewable Energy


Thank you

Questions?

Renewable Industries: From Blade to Grid Ian Jones* … UK/staticassets... · Renewable Industries: From Blade to Grid ... Inc. Proprietary Velocity Magnitude Case Studies: ... –Unsteady

Documents