SUPERGEN Wind

CFD Modelling of Wind Flow In And Around Forested RegionsForested Regions.

Scott WylieScott Wylie Supervised By : Dr Simon Watson

Presentation OutlinePresentation Outline

Supergen Wind Introduction to my research Wind characteristics and Resource Assessment

General concepts The ABL

Forests Past work Basic Theory

Computational Fluid Dynamics (CFD) What is it? Forests and CFD

Current and Future work

SUPERGEN WindSUPERGEN Wind

EPSRC Flagship, Est. March 2006 as part of the S t i bl P G ti d S lSustainable Power Generation and Supply (Supergen) programme

The Consortium is divided into 4 separate themes of The Consortium is divided into 4 separate themes of research:

Theme W: Baselining Wind Turbine Performanceg Theme X: Drive-train Loads & Monitoring Theme Y: Structural Loads & Materials Theme Z: Environmental Issuese e o e a ssues

"To undertake research to improve the cost-effective li bilit d il bilit f i ti d f t lreliability and availability of existing and future large-

scale wind turbine systems in the UK."

My ResearchMy Research

Using CFD, Ansys CFX 11, to simulate effects f t l t i tof two complex terrain types

Forests Following on from work done by John Walshe

(2003) Offshore Looking at the thermal stratification and itsLooking at the thermal stratification and its

effect on wind flow

Justification For ResearchJustification For Research

S it bl h it l ti b i Suitable, onshore, site selection becoming increasingly difficult

So, more remote / less desirable sites being considered – forests being one of these!

To be exploited to their full potential accurate prediction tools are necessaryp y

Resource assessment is one of, if not the, most important factor when applying formost important factor when applying for planning permission.

Wind Characteristics and Resource AssessmentWind Characteristics and Resource Assessment

Gl b l i d tt R h T ( ) Global wind patterns Coriolis Force (or effect)

G t hi Wi d Z

Roughness Type zo (m)

Sea, Sand, and Snow ~ 0.0002

Concrete, Dessert, and Fl t tid

0.0002 – 0.0005

Wind speed over Electronic and Electrical Engineering building (24/04 to 08/05)

Geostrophic Wind Variable nature of wind x

ZfgVg

Flat tides

Flat snow field 0.0001 – 0.0007

Rough ice field 0.001 – 0.012

10

12 Terrain type Two main types of prediction

th d Zg

xf Fallow ground 0.001 – 0.004

Short Grass 0.008 – 0.03

Long grass and Heather 0 02 0 06

6

8

Win

d sp

eed

(m/s

)methods Statistical Physical y

ZfgUg

Long grass and Heather 0.02 – 0.06

Low crops 0.04 – 0.09

High crops 0 12 0 18

0

2

4WPhysical yfHigh crops 0.12 – 0.18

Pine forest 0.8 – 1.6

Town 0.7 – 1.51 991 1981 2971 3961 4951 5941 6931 7921 8911 9901 10891 11881 12871 13861 14851 15841 16831 17821 18811 19801 20791

Number of smaples (1 minute averages)

The Boundary LayerThe Boundary Layer

ss

o Lz

zz

kU

zU ln)( * Stability Function

ForestsForests

A lot of work was done in the 80’s and 90’s A lot of work was done in the 80 s and 90 s Conclusions from this work indicated that

dd d l it d i d t b ladded complexity and increased turbulence made them unsuitable for turbine installation

Complexity of sites (remoteness, ease of access) makes detailed wind field studies difficult

Advances in CFD and computing power mean p g pmore detailed research can be done to give better ideas about turbine placement.p

Forests continuedForests continued

Height 1 a is the forest floorHeight

d

e

2 – 3h

1. a is the forest floor2. b is the sub-canopy3. c is the canopy4 d is the roughness

c

d

z = d +z

h

4. d is the roughness sublayer

5. e is the surface layer

a

b

cz d +zo

U

Wind speed

a

dzu *

ss

o Lz

zz

kU

zU ln)( *

zodzu

zu ln)( *

What is CFD?What is CFD?

Pre-Processor

Solver Post-ProcessorProcessor Processor

A computational technology that enables you to study thethat enables you to study the

dynamics of things that fl

•Defining the geometry for the region of interest: the computational DOMAIN.G id ti th b di i i f th d i i t b f ll

•Approximations of the unknown flow variables by means of simple functions.•Domain geometry and grid display•Vector plots

flow.•Grid generation - the sub division of the domain into a number of smaller, non-overlapping sub-domains: a grid (or mesh) of cells (or controls volumes or elements).

•Discretisation by substitution of the approximations into the governing flow equations and subsequent mathematical manipulations.•Solutions of the algebraic equations (could be user defined)

•Line and shaded contours •2D and 3D surface plots•Particle tracking•Selection of the physical and chemical phenomena that needs to be modelled.•Defining the fluid properties.•Specification of appropriate boundary conditions at cells which coincide with

Solutions of the algebraic equations (could be user defined)Particle tracking•View manipulation•Colour post script output.Specification of appropriate boundary conditions at cells which coincide with

or touch the domain boundary.

CFD Implementation Of A ForestCFD Implementation Of A Forest

Logarithmic wind profile with developed boundary l ilayer as input

Forest canopy modelled as a resistive porous volume (single/multiple), which can absorb momentum

Inclusion of source/sink terms may be useful Investigation in to forest clear cuts also of interestInvestigation in to forest clear cuts also of interest.

Log profile at inlet, developed boundary

Log profile at inlet, developed boundary p y

layerp y

layer

Individual trees modelled, rather than single porous volume

Porous volume, possibly layered to account for different density at different levels

Initial CFD Attempts: What it Looks LikeInitial CFD Attempts: What it Looks Like

PlotsPlots

Horizontal ChangesMiddle of Forest Comparison

70

801.75

2

50

60

70

l [m

]1.25

1.5

20

30

40

Hei

ght a

gl

0.75

1 chz

0

10

0 2 4 6 8 10 12

Wind Speed [m/s]0.25

0.5roughness lengthresistive volume

Vertical ChangesUniform 1 / m 1.25 / m 1.5 / m

0 0.25 0.5 0.75 1 1.25 1.5 1.750

hcuu

Current WorkCurrent Work

EWEC 2009 Marseille 16th 19th March: EWEC 2009, Marseille 16th – 19th March: Poster Presentation

CFD Modelling and Validation of Wind Flow CFD Modelling and Validation of Wind Flow over A Forest Canopy

ASME 3rd Int. Conference on Energy Sustainability San Francisco 19th 30th July:Sustainability, San Francisco 19th – 30th July: Abstract excepted

A C t ti l Fl id D i St d f Wi d A Computational Fluid Dynamics Study of Wind Flow Across A Forest Edge

EWEC 2009EWEC 2009CFD Modelling and Validation of Wind Flow over A

Forest CanopyForest Canopy CFD study of the wind flow over a small forested

region Hill Holt Wood in Lincolnshire UKregion, Hill Holt Wood, in Lincolnshire, UK. Single mast at site, measuring wind speeds at two

different heights (11.3 m & 3.2 m)different heights (11.3 m & 3.2 m)

ASME 3rd Int Conference on Energy SustainabilityASME 3rd Int. Conference on Energy SustainabilityA Computational Fluid Dynamics Study of Wind Flow

Across A Forest EdgeAcross A Forest Edge

CFD comparison with Wind tunnel data simulating5

CFD comparison with Wind tunnel data simulating Harwood Forest, Northumberland.

Experiment carried out at Oxford University4

Experiment carried out at Oxford University environmental wind tunnel.

1/75 scale was used; 104 x 104 trees spaced at

3

z/h

x/h = 3.5

x/h = -11/75 scale was used; 104 x 104 trees spaced at 0.0231 m, each being 0.2 m high

1

2 x/h = 10.5

00 0.5 1 1.5

U/Uref

Future WorkFuture Work

Continue to improve accuracy of forest CFD models through validation of field and wind tunnel data

Work with Surrey University – Looking at physical y y g p y(wind tunnel) modelling of forests

Begin to look at modelling near offshore region?Log profile at inlet, developed boundary layer

Land region, heat source Sea Region, different heat dissipation rate compared to that of land region.

To ConcludeTo Conclude…

Initial modelling attempts have highlighted h i t d d!areas where improvements are needed!

Data acquired needs detailed analysis Hill Holt Harwood Forest

D t il d t d f t d l i d d Detailed study of past models is needed John Walshe’s attempts at modelling forests

M t d W t (f ff h ) Montavon and Watson (for offshore) Conference papers will allow for some more

detailed simulations to be ran and analyseddetailed simulations to be ran and analysed Some initial modelling of land/sea interface are

needed should time permitneeded, should time permit.

Any Questions?Any Questions?

Thank you for your time!