Teaching wind resources, wind energy planning and projects ...€¦ · 3 November 2016 This course has been put together for the purposes of teaching planners, assessors and students

Wind Atlas for South Africa project

Teaching wind resources, wind energy planning and projects

Presentation of “A training course in applying the products of the WASA project”

Tom CroninDTU Wind Energy

Wind Atlas for South Africa projectWind Atlas for South Africa project

Conditions of use for this training material

3 November 2016

This course has been put together for the purposes of teaching planners, assessors and students about wind atlases and, specifically, the Wind Atlas for South Africa. All users are asked to note and respect the following:

• Individual slides must be used in their entirety.

• Any use of the material must be given the appropriate reference: “Wind Atlas for South Africa Training Material May 2014”.

• Separate topics may be presented individually (e.g. “What is a Wind Atlas?”).

• The course is, however, intended to be given as a complete course by suitably qualified persons.

• For an introduction as to how the course is meant to be given, please refer to the video: “WASA sample of video notes accompanying the training material 4June 2014”

• All non-original graphics, graphs, images, etc. are referenced either on the slide itself or in the notes for each slide.

• Graphics, graphs, images etc., that are shown without references are the result of original work in the WASA project.

This material is not for public distribution. For users other than the WASA project partners, prior permission must be obtained

from SANEDI before using this material

WP4 Application Training Course2


The Wind Atlas for South Africa: WASA

A training course in applying the products of the WASA project

PlanningReviewing potential wind farm productionEducation

A product of the WASA project: a wind resource map of Overstrand overlaid onto a Google Earth image


Teaching structure of this course material

3 November 2016

Theme 1e.g. wind atlases 1A 1C1B

This training course is built up of ‘themes’ that are related….This training course is built up of ‘themes’ that are related….

… and each theme has a track… and each theme has a track

Theme 2e.g. the wind

industry sector2A 2B

Theme 3e.g. WASA project 1A 1C1B

Teaching notes

Course structure Teaching methods Target audience



Teaching structure of this course material

3 November 2016

Theme 1e.g. wind atlases 1A 1C1B

Theme 2e.g. the wind

industry sector2A 2B

2CTheme 3e.g. WASA project 1A 1C1B

The idea is that the themes can be moved or swapped…The idea is that the themes can be moved or swapped…

…and the tracks can be more detailed or simplified… …and the tracks can be more detailed or simplified…

…to create the course that suits the target audience.

…to create the course that suits the target audience.

Teaching notes




Teaching methods for this course material

The techniques to be used for this course are designed to be varied and include:

• Presentations• Interactive sessions with the audience• Question & Answer sessions• Case studies (more or less ‘hands on’)• Videos

3 November 2016

Teaching notes




Target audience for this course

This course material has been put together for:

• Planners• Local government employees• Students of wind energy

It is not especially suited for• Developers• Consultants• Manufacturers

Nor is it a wind resource assessment course; but it is for everyone who wants to know what a wind atlas is and how to use one.

3 November 2016

Teaching notes






PlannersLocal authoritiesStudents


Introduction


Training course expectations

3 November 2016

Interaction

Let’s hear your expectations:• Who are you?• What do you do?• What do you expect from this course?

My expectations:• You will ask questions

(I cannot promise to know all the answers but I bet I’ll be able to find someone who can answer your question!)

• You will involve situations from your everyday work.• Participate and discuss (with me and your fellow participants)

Introduction Expectations Objectives Course outline



Course objectives

3 November 2016

After this course, you should know:

• What a wind atlas is• How a wind atlas is made• Why a wind atlas is so useful• How the results from the WASA project can be used to

– Help in planning for wind energy– Help when reviewing the estimated production from a wind farm– Learn more about the wind energy resource

And finally:

• A better idea of how the WASA can help you in your job• Where to go for more information




Course outline

3 November 2016

• What is a wind atlas?• Who needs a wind atlas?• The WASA project• Wind resource assessment for wind energy• What’s involved in a good wind farm project?• What can a wind atlas actually be used for?• How is a state-of-the-art wind atlas made?• Accessing and using the WASA data• Case studies• Good practice use of data• Wrap-up and possible next steps




Course outline

3 November 2016





What is a Wind Atlas?

3 November 2016

What is a wind atlas? Contents Properties Example

What do you think?

• What is a ‘normal’ atlas?• What type of information does an atlas give?• What can they be used for?• ‘Atlas’ versus ‘map’: what’s the difference?



What does a wind atlas contain?

3 November 2016

What is a wind atlas? Contents Properties Example

A wind atlas is more than a coloured picture…

• A database of wind information i.e. numbers• Wind speeds• Wind directions• Frequency of wind conditions• …and can also be a graphical representation i.e. a coloured picture



An example wind atlas picture

3 November 2016

What is a wind atlas? Contents Example Properties



Properties of a wind atlas

3 November 2016

What is a wind atlas? Contents Example Properties

Just like any atlas, a wind atlas has specific conditions for the information

• Wind climate statistics• Terrain has been reduced to a flat plane• Uniform land cover• Specific heights above the ground

For the WASA wind information:

• Annual mean wind speed, direction distribution and speed frequency

• 100m above ground level (a.g.l.)

• Flat terrain

• 3cm ‘roughness length’ everywhere



Course outline

3 November 2016

• What is a wind atlas?• Who needs a wind atlas?• The WASA project• Wind resource assessment for wind energy• What’s involved in a good wind farm project?• What can a wind atlas be used for?• How is a state-of-the-art wind atlas made?• Accessing and using the WASA data• Case studies• Good practice use of data• Wrap-up and possible next steps



The wind energy sector

3 November 2016

Who needs a wind atlas?(or…who needs wind data?)

Who? Brainstorm Sector players Uses & users Conclusion




3 November 2016

Brainstorming session:

Given that the Government of South Africa has embarked on a programme for providing the nation with increased wind power, who needs wind information?





3 November 2016

Local planning authorities

Wind energy consultants




Users and uses

3 November 2016


Authorities Policies, regulations and plans

Planners Resource and development planning

Investors, owners and banks Financial planning, risk assessment and decisions

Developers (small and large) Project development

Industry (small and large) Project design and implementation,Wind turbine design and development

Power sector Power system planning, development and operation

Consultants Independent expertise and tools development

Academic community Research, tools development & education



Conclusion: who needs a wind atlas?

3 November 2016


Nearly anyone connected with the wind energy sector can benefit from using the wind atlas project results for (but not limited to):

• Geographical planning of activities• Estimation of wind energy production• Education and research

Just about the only situation where a wind atlas cannot (yet) be solely used is in producing bankable wind resource predictions for wind farms.



Course outline

3 November 2016




The WASA project

3 November 2016

Producing much more than ‘just’ a wind atlas

The WASA project Project results Domain area Project team



WASA project products (results)

3 November 2016

All products of the WASA project are in the public domain:

• Verified Numerical Wind Atlas (×2)• 10 high quality wind data recording masts• Time series data from the masts• Local wind atlases at the masts• Wind resource mapping around the masts• High resolution wind resource maps of the whole domain• Extreme wind atlas• Wind forecasting• Power production forecasts• Reports, papers, presentations, etc.• Guides, best practice documents and workshop training

Finally, but not least:• Websites and software for downloading all the above!The WASA project Project results Domain area Project team

We’ll learn about all

these as we go through the course!



WASA project and domain area

3 November 2016

The WASA project started in June 2009 and finished in April 2014

The domain covers:

• The Western Cape

• Part of the Northern Cape

• Part of the Eastern Cape

In total: 350 000 km2 of RSA (~29%)




The WASA project drivers and partners

3 November 2016

The project is an initiative of the South African Government - Department of Energy and co-funded by:• UNDP-GEF through the South African Wind Energy Programme (SAWEP)• Royal Danish EmbassyProject Steering Committee: • DoE (chair), DEA, DST, UNDP, Danish Embassy, SANEDIExecuted by:• The South African National Energy Development Institute (SANEDI) Implemented by: • CSIR, UCT, SAWS, and DTU Wind Energy (formerly Risø DTU).




Course outline

3 November 2016




Wind resource assessment

3 November 2016

Wind resource assessment

and The Wind Atlas Methodology

Introduction Wind resource Wind atlas method Traditional method State‐of‐the‐art



Wind resource assessment: why?

3 November 2016

Why do we need to know?

(May be obvious but it’s important to consider)

• Planning of wind farm areas: political and environmental• Estimation of wind farm power production• Power system - long term and large investments• Value of location of best wind resources• Temporal and spatial wind power distributions The wind is not the same everywhere, all the time




Wind resource assessment: wind data

3 November 2016


From wind measurements (‘observations’)…

…to wind statistics (‘climate’)

Frequency of direction (wind rose)

Frequency of speed(histogram)Wind atlas method



Wind resource assessment: the challenge

3 November 2016

The wind resource challenge:

Physically measuring the wind is only valid for that location and time period

• It is expensive to measure the wind if we are not certain we can build a viable wind farm there

• It is expensive and impractical to measure everywhere we think we need to makes plans for

So, how important is it that measured data can be ‘adjusted’ so it is appropriate for another location?




Wind resource assessment: the importance

3 November 2016

Wind provides the income in cost-benefit

Investment costs Operation and maintenance costs Electricity production ~ Wind resources Turbine lifetime Discount rate Environmental benefits

Modelling is necessary and it must be good

Energy in the windP = ½U3 [W/ m2]

Energy in the windP = ½U3 [W/ m2]

Wind speed U [m/s]

Wind speed U [m/s]

10% error on speed up to 30% error on energy




Measurements + Linear interpolation = NO

3 November 2016




Wind Atlas Methodology: uses flow modelling

3 November 2016

• Analysis procedure ( )Observed Wind Climate

+ sheltering obstacles+ roughness map+ elevation map

→ Generalised Wind Climate

• Application procedure ( )Generalised Wind Climate

+ sheltering obstacles+ roughness map+ elevation map

→ Predicted Wind Climate

“Observed” = measured i.e. from a meteorological mast

“Predicted” = modelled at another place. Can be used

for predicted turbine production




Application 1: prediction of energy production

3 November 2016

Observed data from one (or more) site

met masts

MO

DELLIN

G A

RO

UN

D TH

E PRED

ICTO

R SITES

Predicted Wind Climate+ power and thrust

curves+ wind farm layout

Predicted wind farm Annual Energy

Production

This prediction relies on:

a) Quality datab) Similar terrain for

observed & predictor sites

c) Observed and predictor sites must have the same generalised wind climate

=> Restriction on how far apart the two sites can be

Introduction Wind resource Wind atlas method Applications State‐of‐the‐art

MO

DEL

LIN

G A

RO

UN

D T

HE

MET

STA

TIO

NS

This is what the majority of wind farm project developers do to be able to obtain financing for their wind farm



Application 2: a ‘traditional’ wind atlas

3 November 2016

Observed data from over 200 met. stations

in Europe

MO

DEL

LIN

G A

RO

UN

D T

HE

MET

STA

TIO

NS

Collection of Generalised Wind Climates

= Wind Atlas

The observational wind atlas method and the microscale flow model, WAsP, were conceived in the 1980’s for the European Wind Atlas (issued 1989)





3 November 2016


in Europe

MO

DEL

LIN

G A

RO

UN

D T

HE

MET

STA

TIO

NS


= Wind Atlas Things to note about theobservational wind atlas:• Influence of obstacles,

real land roughness & terrain removed

• Resulting wind data is statistical (not time series)

• Wind data is given at 50m a.g.l.

• Modelling is at microscale (‘local’ features are modelled e.g. towns, forests, hills, valleys, etc.)





3 November 2016


in Europe

MO

DEL

LIN

G A

RO

UN

D T

HE

MET

STA

TIO

NS


= Wind Atlas Drawbacks of the observational wind atlas:• Output resolution rather

coarse

• Requires many met. masts

• Most masts were not designed for wind energy purposes

• If many ‘wind energy’ met. masts were to be used: Difficult Expensive Time consuming





3 November 2016

With increasing computer power in the 1990’s, a new way was developed of obtaining generalised wind climates using weather models:• Wind data not from ground-

based measurements…

• …but large atmospheric simulations…

• …and ‘mesoscale’ models (regional features)

• Producing a numerical wind atlas Large domains Long-term wind climates Time-efficient Less expensive

Wind data from atmospheric weather

models

And this is what the state-of-the-art WASA wind atlases are based on

‘modern’


MES

OSC

ALE

MO

DEL

LIN

G O

F LA

ND

SCA

PE



Wind Atlas: a summary of concepts so far

3 November 2016

• Time series measurements provide statistics about a wind climate • Remove local effects of terrain to provide Generalised Wind Climates• Using modelling of local terrain the GWC can be used to provide a

Predicted Wind Climate in another location (not too far away!)• Local terrain effects are modelled at micro scale• A traditional Wind Atlas is a collection of GWCs based on measurements:

an observational wind atlas• A modern Wind Atlas uses wind data from weather simulations to cover

large regions: a numerical wind atlas• Regional terrain effects are modelled at meso scale

“The state-of-the-art for wind resource assessment and planning is a combination of microscale and mesoscale

modelling using measurements for verification”

Introduction Wind resource Wind atlas method Applications Concepts ‐ summary



Wind resource assessment & wind atlases

3 November 2016

The WASA wind atlases represent this state-of-the-art…

…which we will look at in a while…

…but now we’re going to take a brief look at:

What makes a good wind farm project?

Introduction Wind resource Wind atlas method Applications Concepts ‐ summary



Course outline

3 November 2016




What’s involved in a good wind farm project?

3 November 2016

EXAM QUESTION #1:

“Sustainable, competitive, socially responsible, clean and useable power from wind relies on good wind farm projects”

Discuss.



Standards, guidelines and good practice• Not (yet) one standard, recognised, way to build & operate a wind farm• Although procedures are converging there are still many different

practices being used by various developers…

• …but the environment for good wind farm projects is set by clear policies and good planning

Main standard in wind energy: IEC 61400 series wind turbine design Turbine ‘classes’: IEC 61400-1 Turbine certification in accordance with IEC 61400-22 Project certification, based on IEC 61400-22 Wind measurement: Measnet and IEA procedures Future standard: Assessment of site specific wind conditions IEC

61400-15 Technical due diligence can be required by investors

3 November 2016WP4 Application Training Course45


The 4 phases of a wind farm project

3 November 2016

Development Implementation Operation Decommissioning



Possible timescales

Development Implementation Operation Decommissioning

3 November 2016

• Finding a site• Wind resource assessment• Wind farm design• Permissions & licences• Financing• Contract negotiations

• Procurement• Construction • Operation

• Maintenance

• Re-power/• Remove

Up to 10 years 1-2 yrs 20 – 25 years <1

Relatively small amount of money but over a long time

and with some risk

Large amount

of moneyless risk

Regular revenue, regular costs, risk reducing further

Small lump sum



The 5 cornerstones of wind farm development

• Energy production• Turbine selection & sitingWind resource

• Power delivery• Energy sales

Grid connection

• Costs• Revenue

Project economics

• Environmental assessment• Public acceptanceSociety

• Permission to build• Stable political policy

Planning and Policy

3 November 2016

Win

d fa

rm p

roje

ct

deve

lopm

ent

(ite

rativ

e pr

oces

s)



1. Wind resource assessment and siting

3 November 2016

Aim: • To find a suitable site, optimise production and visual impact whilst

ensuring design conditions are met, estimate energy yield and assess the uncertainty of this

Aspects:• Prioritise areas of good wind resource• Select candidate sites• Carry out wind measurement campaign(s)• Analyse resource and terrain• Site turbines• Advise on turbine class required• Estimate energy yield• Make uncertainty assessment



IEC 61400-1 site assessment rules for turbine class required• Extreme winds• Vertical wind shear• Flow inclination• Turbulence• Wind-speed distribution

1. Wind resource assessment and siting



2. Grid connection

3 November 2016

Aim: • To develop a grid connection that can deliver the wind farm power to the

grid, be cost effective and meet current regulations

Aspects:• Part of site selection process: sites can be

ruled out because grid connection is too expensive

• Find grid connection point with suitable strength for max wind farm output

• Select suitable overhead lines and/or cabling and comply with planning regulations

• Design internal grid layout for wind farm• Ensure compliance with grid codes• Develop suitable power purchase

agreement



3. Project economics

3 November 2016

Aim: • To demonstrate that the project is sufficiently profitable for the developer

and for the investors

Aspects:• Part of the site selection process• Revenue comes from the wind resource• Establish key figures for the wind farm Price of selling electricity Capital required for building wind farm Operations and maintenance costs Decommissioning costs Establish economic factors (inflation

rates, exchange rates, etc.) NPV and IRR of the project• Obtain financing



4. Society: Environment and public acceptance

3 November 2016

Aim: • To ensure the impact of the wind farm on its environs (both human

population and natural surroundings) is in balance with the generation of power

Aspects:• Part of the site selection process: some

sites are ruled out due to adverse impact on the environmental or the public

• Legislation for environmental impact and public consultations should be followed

• Carry out environmental surveys and studies to produce an Environmental Impact Assessment (E.I.A)

• Adjust the wind farm design accordingly• Inform the public and involve them as

part of the project at the earliest stage• Note: power generation will never be

impact-free

Ansøgning til regionplanmyndighed

Møller > 80m Antal > 3 Screening Ikke VVM pligtig

Offentliggørelsemed

klagevejledning

VVM pligt

Idéfase VVM

redegørelse

Vedtagelse af regionplan,

kommuneplan og lokalplan

Offentlig høringmin 8 uger

Behandling af indkomne forslag

Offentliggørelse

med klagevejledning

VVM tilladelse

og byggetilladelse



5. Planning and policy

3 November 2016

Aim: • To obtain the permission and permits necessary for building and

operating the wind farm

Aspects:• Part of the site selection process• With good planning and zoning, done on

the basis of reliable data, suitable areas can be quickly identified

• The development of good and cost efficient wind farm projects will thrive where the planning and approval process is straightforward and transparent

• This starts with a clear policy set at the national level, allowing effective implementation of the policy at local level.



Conclusion: wind farm project development

3 November 2016

Development of good wind farm projects:

• Is a complex process• Needs to be good: wind farm construction is capital intensive• Involves many types of specialists to be co-ordinated• Relies on good quality data• Depends heavily on the right policy and planning framework• …

…which is all very difficult to get right if we do not know accurately where the good wind resource is!

So this brings us back to the wind atlas



Course outline

3 November 2016




What can a wind atlas actually be used for?

Three brief examples

(More details demonstrated later with the case studies)

1) Input to national policy-making processes

2) Planning and zoning at a regional level

3) Wind farm project planning



Example 1: National policy

Governmental policy• A wind atlas can show the potential of wind power in a country• Allows policy for wind power support to be co-ordinated with e.g.

employment policy, industry and commerce policy, etc.• Important input to any support mechanism decision (e.g. feed-in tariff,

bidding process, market price plus, etc.)• Helps to shape realistic targets for policy goals and interval steps for

getting there

National grid network development policy• Input to grid expansion and strengthening strategies• Helps make large national investments more cost-effective

3 November 2016

2 GW by

2020



Example 1: National policy

3 November 2016

Reference: DEA Strategic Environmental Assessment by CSIR, source: http://www.csir.co.za/nationalwindsolarsea/images/2_Transmission%20Corridors_2.jpg



Example 2: Regional planning & zoning

Making development zones for co-ordinated planning1

• CSIR, on behalf of the DEA, have worked with the WASA data as part of the Wind and Solar PV Strategic Environmental Assessment study

• “To facilitate the efficient and effective rollout of wind and solar PV energy in South Africa”

• Objectives: Identify Renewable Energy Development Zones (REDZs) for large scale

projects Minimise adverse impact on the environment whilst maximising benefit

for society Decrease the risk of wind farm development within REDZs by

screening and co-ordinated planning in advance and streamlining the environmental authorisation process

3 November 2016

1 Reference: SEA Progress and Study Areas, presentation Windaba 2013, Cornelius van der Westhuizen, CSIR




3 November 2016

Conditioning of WASA data •Screen for areas of good wind potential

Mapping of development potential •Consider aspects such as proximity to ports, employment, etc.

Mapping of no-go areas • Exclude areas e.g. environmentally sensitive

Identification of study areas

Reference: SEA Progress and Study Areas, presentation Windaba 2013, Cornelius van der Westhuizen, CSIR




3 November 2016

Reference: SEA Progress and Study Areas, presentation Windaba 2013, Cornelius van der Westhuizen, CSIR



Example 3: Wind farm project planning

A developer has identified a potentially good area for a wind farm

• Need to make a wind measurement campaign for a bankable project• Exact location of wind farm still uncertain – where to measure?• The wind atlas can quickly help in the decision of how many masts and

where they should be located so as to ensure the data is reliable enough to be used over a suitable area to cover possible site(s)

• It is important to get this right because: Measuring in the wrong place is expensive The measurement campaign takes time (>1 year) just “re-doing” the

measurements is not an option



Example 3: Wind farm project planning



Course outline

3 November 2016




Summary of the course so far

3 November 2016

So far we

have an idea of what a wind atlas is and what it contains

know who can make use of a wind atlas

have looked briefly at the WASA project

found out why knowing the wind resource is important

outlined what is involved in wind farm development

and seen examples of what a wind atlas can be used for.



Summary of the course so far

3 November 2016

So we now need to

look at how a state-of-the-art wind atlas is made

So we can

better understand how to use it

use it fully in combination with other information

and be aware of the accuracy and limitations of the data.



How is a state-of-the-art wind atlas made?

We looked at the ‘traditional’ observational wind atlas.• What was the main challenge for producing an atlas with better resolution?

• The types of methods and processes used in producing a state-of-the-art wind atlas all depend on the source of the wind data

We then briefly introduced the numerical wind atlas.• The wind data used for this ‘modern’ type of atlas comes from simulations

of the atmosphere (i.e. weather models)

3 November 2016

Think back to the section on wind resource…

Good quality wind data over a large enough area and time period

Let’s start with these weather models…..




3 November 2016

• We need to review the concepts of modelling scales because

Model resolution Data input Data output Purpose of the model

…are all linked together and are decisive in using the data correctly!

For instance, modelling the wind using a coarse resolution (=small scale) using terrain information at a coarse resolution will not give the same results as modelling the wind at a fine resolution (=large scale) using terrain information at a fine resolution.

Which is one good reason why a wind atlas data is given for winds over a flat terrain with a constant roughness

…but just before we look at weather models…



Modelling scales

• Here are the scales that you’ll hear used very often

3 November 2016

• MACRO-scale• Global• grid size ~200×200km

• MESO-scale• Regional• grid size ~5×5km

• MICRO-scale• Local• grid size ~50×50m

(“reality” included not just as a joke but a reminder that all models are approximations of reality)

• Actual• Infinitesimal



Modelling scales

• Here are the scales that you’ll hear used very often

3 November 2016

• Macroscale• Global Circulation Model• Medium-range weather

forecasting

• Mesoscale• KAMM/WRF/MM5…• Wind atlas

• Microscale• WAsP/MS-Micro/CFD• Wind farm production

estimation• Wind energy density

maps

• Actual• The wind you

feel on your face

The process of going from a large scale to a smaller one is “downscaling”

Example wind-related purposes:

Getting the interfaces right between the modelling scales and downscaling correctly is one of the major challenges of making a wind atlas



Modelling scales and making a wind atlas

• By now, you can probably guess the main steps in making/using a wind atlas…

3 November 2016

output

Global weather

data

Mesoscale terrain

inpu

t

output

Simulated and generalised winds

Numerical Wind Atlas

Microscale terrain

inpu

t

output

Win

d sp

eed

and

pow

er d

ensi

ty m

aps

Win

d fa

rm e

nerg

yPr

oduc

tion

estim

ates

…and we can now start with weather models!



Wind data input from weather models

3 November 2016

Global Circulation Models (GCMs)

Annually averaged winds across the world

There are now many sources of these “reanalysis” data

Reanalysis = weather model is re-run using all the data collated over the last 30-40 years to arrive at the data set

Source: European Center for Medium Range Weather Forecasting (ECMWF)ERA Interim reanalysis

So why not just use this data?

Remember what we said earlier about scales and purpose?



Wind data input from weather models

The general patterns shown by the GMC models do not capture regional phenomena that are at the meso scale

3 November 2016

sea-land breeze coastal jet

gap flowmountain-valley breeze

Therefore we need to downscale the data to be able to capture these effectsWP4 Application Training Course74


Downscaling using mesoscale modelling

The WASA project has, in fact, used two modelling platforms to downscale

• They use slightly different inputs • Work in different ways • Both lead to a generalised wind atlas data set

So, the WASA project has produced two numerical wind atlases.

3 November 2016

Actually, there are three, when we include the extreme wind atlas…but more about that later

Numerical wind atlas

Mesoscale modelling

Generalised climatological wind climates @ grid points



Downscaling using mesoscale modelling

Question: Why use two modelling platforms?

Answer: The first one provides benchmark results and builds confidence The second advances the state-of-the art

Platform 1: Karlsruhe Atmospheric Mesoscale Model (KAMM)• Chosen to be the interim WASA output: “the benchmark”• Has been in use since the 1990s • Industry standard: gives good results and is relatively quick• However, some influential effects not modelled and known to

generally underestimate the wind resource

Platform 2: Weather Research & Forecasting model (WRF) (pronounced “warf”)

• Used to produce the final definitive state-of-the art wind atlas• ‘Next-generation’ model emerging from latest research• Allows modelling of more complex thermal effects• Uses significant computing resources but potentially gives excellent

results



Downscaling: the modelling process

3 November 2016

Statistical macroscale wind input

First numerical wind atlas (2012)

Mesoscale terrain model 5×5km grid

Simulation of many wind situations over the mesoscale terrain using the

KAMM model

Adjustment and treatment of data to get generalised wind climates,

independent of the mesoscale terrain

Generalised wind climates at 15000 points 5×5km grid = Wind Atlas 1• Flat terrain• Constant surface roughness• Specific heights a.g.l.

Time series macroscale wind input

Second numerical wind atlas (2014)

Mesoscale terrain model 3×3km grid

Simulations of the wind over time at each model grid point using the WRF

model

Adjustment and treatment of data to get generalised wind climates,

independent of the mesoscale terrain

Generalised wind climates at 40000 points 3×3km grid = Wind Atlas 2• Flat terrain• Constant surface roughness• Specific heights a.g.l.

Inpu

tSim

ulat

ion

Trea

tmen

tO

utpu

t



What does the WASA output look like? (1)

• Each grid point has an associated generalised wind climate• The wind climate is described in a file suitable for further use in other

analysis software e.g. WAsP.

3 November 2016

• Each grid point is identified by its co-ordinates

• The data can be downloaded either individually for each point or as a whole set in a zip file

• The website for download can be accessed by anyone (after registration)

• Address: http://wasadata.csir.co.za



WASA output download interface: Tadpole

• The wasadata.csir.co.za website gives access to the Tadpole interface

3 November 2016

We’ll see just how to download later on but there’s also another output...



What does the WASA output look like? (2)

• If the annual mean wind speeds from all the generalised wind climates are plotted on a map and given a colour scale then it looks like this:

3 November 2016

m/s

Generalised climatological (30 year) annual mean wind speed, 100m above ground level over flat terrain with a constant surface roughness length of 3cm everywhere

The wind atlas map

or generalised wind map

Please note the accompanying description!



Summary of making a wind atlas (so far)

• Use global, macroscale, wind data from Global Circulation Models• Downscale by simulating the wind climate over the domain terrain

modelled at mesoscale• Obtain the generalised wind climates by converting the terrain to flat

terrain with a constant surface roughness• The result is a grid thousands of data sets, more easily visualised by

plotting appropriate values using a colour scale to get a wind atlas map

3 November 2016

But the question some of you are wondering about is:

If the whole process is based on wind input from simulations then just how good are the results?

And perhaps the rest of you are thinking:

There was something about wind measurement masts in the WASA project. Why bother with that if everything is modelled?

Maybe unsurprisingly, these two are related



The verified numerical wind atlas

• A real state-of-the-art wind atlas is verified by measurements (not all wind atlases are)

• The Wind Atlas for South Africa project was designed from the beginning to include high quality measurements against which the numerical wind atlas could be checked

• This produces a “Verified Numerical Wind Atlas”

3 November 2016

So, alongside the mesoscale modelling, the WASA project has a second, parallel, activity:


Mesoscale modelling


High quality measurements

Microscale modelling

Generalised wind climates @ mast locations Ve

rific

atio

n



The WASA wind measurement campaign

• 10 meteorological masts were installed

• Positions chosen carefully to cover the WASA domain be in the representative

terrain types and climatology found in the domain

be suitable for modelling• Not chosen for the ‘windiest’ locations or ‘best locations’ for wind

farms• Data automatically recorded and

downloaded remotely• Designed and installed to IEC

standards and Measnetguidelines



Features of high quality wind measurements

• Instrumentation arranged to minimise errors and uncertainties due to flow distortion

• Proven sensors of high quality and individually calibrated

• Regular maintenance• Anemometers at five different heights: 10,

20, 40, 60 and 62m• Wind vanes at two heights: 20 & 60m• Temperature and pressure sensors• Data recorded at 10 minute average

intervals• Turbulence calculated and recorded• Data downloaded and checked regularly to

minimize loss of data or data gaps

3 November 2016

WM9 Noupoort 60-m lattice mast



A great job…on a good day



Using the measured data

• The data is first quality checked by the WASA team• Then all the data is uploaded to the same website as the numerical wind

atlas data: http://wasadata.csir.co.za• All data available free to the public for use at their own risk• Every mast online since mid 2010• Data is in csv format (monthly)

3 November 2016

We’ll see just how to download later on but how was it used for verification?



Verification of the wind atlas by measurements

• The local terrain around each mast was modelled at micro scale

• Elevation and roughness maps created of 20km × 20km

• 10m contour lines• The measured wind data was

converted into 10 Observed Wind Climates (OWC)

• The WAsP microscale model then used to obtain the Generalised Wind Climate (GWC) for each mast area

3 November 2016

Remember what we said earlier about scales and purpose?

MO

DEL

LIN

G A

RO

UN

D T

HE

MET

STA

TIO

NS

So we now have 10 GWCs from the masts and 15000 & 40000 GWCs from the numerical wind atlases



Verification of the wind atlas by measurements

• So we can compare the numerical wind atlas GWC that is closest to each mast with the GWC derived from the mast data

3 November 2016

WM10 GWCNWA GWC

Please note:• Both sets of GWCs

must have the same attributes i.e.

• Same height a.g.l.• Flat terrain• Uniform roughness

(The NWA data was also adjusted so that it was representative over the same period the met mast measurements were taken.)



What are the results of the verification? (1)

• Comparison of the generalised wind climates• Annual mean wind speeds at 100m• Model roughness height of 3cm• Average of comparisons at the 10 mast sites

3 November 2016

Or…just how good are the wind atlases at the

comparison locations?

NWA KAMM NWA WRF

Mean deviation from OWA 6.4% 4.4%

Bias -4.8% 2.1%

On average there is a difference of about 6% (either above or below) between the NWA KAMM results and the observed values. Considering all sites together, there is a

tendency to under-estimate by around 5%.

The WRF model results in a wind atlas with winds speeds both closer to the

measured wind speeds and with a smaller bias; this time a slight over-prediction

This is a significant improvement!



What are the results of the verification? (2)

• Comparison of actual wind speed predictions• Annual mean wind speeds at 80m• Site around masts modelled at microscale

3 November 2016

The WRF-based NWA GWCs nearest to each measurement masts were used to predict the

annual mean wind speeds at the measurement mast positions at

80m (a typical hub height), using microscale modelling. These are

then plotted against the results of doing the same calculations but

using the OWCs as input.

On average, the NWA calculations over-predict the wind speed by

about 3%



It is not all about average wind speeds

• A comparison of the distributions of wind directions wind speeds

• … for the WRF-based NWA also shows very good agreement

3 November 2016

WM05 Observed wind atlas WRF-based NWA



Generalised wind speeds – based on WRF mesoscale modelling mean wind speed [m/s] 100 m above ground level, flat terrain, 3 cm roughness everywhere

The Wind Atlas for South Africa: final wind map

3 November 2016

WRF-based Numerical Wind Atlas (NWA)

March 2014



The Wind Atlas for South Africa: final wind atlas

In summary• The verification tasks have shown that the WRF-based NWA gives

excellent results at the 10 measurement mast locations• Not only are the annual mean wind speeds accurate but the distribution

of the directions and wind speeds are also represented very well• The value of the KAMM-based NWA is in backing-up the WRF-based NWA

in areas where there are no verification masts: the trends are generally well matched.

3 November 2016

The WRF-based wind atlas truly represents the verified state-of-the-art and is the recommended data set for all users

Users must still be aware that, whilst on average the results are good, at any given location there can be over or under estimates

(more on this later at the end of the course)




• We can now look at a third activity in the project: extreme winds

3 November 2016

We can now answer this question because we have look at how the two WASA wind atlases were made and verified…

…but wasn’t there something about a third wind atlas mentioned earlier?


Mesoscale modelling




Generalised wind climates @ mast locations

Extreme winds Mesoscale modelling

Generalised extreme wind climates @ grid points

Veri

ficat

ion



Extreme winds and their importance

• Extreme winds are – qualitatively – strong winds that could damage a turbine but do not occur that often.

• In wind energy, and in the IEC standards, we use the 50 year return wind• A wind speed that is, on average, exceeded only once in 50 years

• Some sites will have a high value for the extreme wind, others a lower value

• This, in part, determines the type of turbine that should be used: Class I, II, III or a ‘special’.

3 November 2016

The challenge is that it is not so easy to measure a wind speed that only comes once in 50 years:We can’t always wait that long and it is difficult to measure!



Determining the extreme wind

• Often where there are wind measurements from a mast: Take the highest wind speed recorded in each year of data Assume that these maxima have a ‘Gumbel’ distribution Extrapolate to find the maximum wind every 50 years

3 November 2016

But we face the same problem as with the wind atlases we have seen: we cannot cover a large area with wind measurement masts to find the extreme winds.



Making an extreme wind atlas

• As with the numerical wind atlas, mesoscale modelling with reanalysis data input is at the centre of producing an extreme wind atlas

• Difference now is that measurement data is also used to identify specific storm events leading to extreme winds

• WRF modelling results are then post-processed to give generalised extreme wind climate data sets

3 November 2016

WASA Phase 1: 50-yr, 10 min wind speed (m/s) @ 10m, z0=5cm• For the extreme wind atlas

there are two domains: Western & Eastern Cape

• All together 27000 grid points at 4x4km resolution

• Data set for each point contains the annual maximum wind speed and direction for each year from 1998 to 2010

• These can be generalised and used to extrapolate the 50 year return extreme wind speed

• Download site:



Outputs from the WASA project

Here’s what has been produced by the WASA project• Two state-of-the-art verified numerical wind atlases• High quality wind measurements from 10 met. masts• Associated observed and generalised wind climates => observational

wind atlas (and WAsP .wwh files)• A state-of-the-art extreme wind atlas

3 November 2016

Numerical wind atlases

Mesoscale modelling





Extreme wind atlas

Mesoscale modelling


Veri

ficat

ion

But that is not all…the microscale modelling has a very significant trick up its sleeve…




• We have the terrain modelled around each of the 10 WASA masts • The GWC has similarly been calculated for each of them• We can now use these to create a resource map at microscale around

each of the 10 masts• Or…model the production from one or a number of turbines in the vicinity

of each of the masts

3 November 2016

Earlier, we learnt that microscale was looking at local effects…



Microscale modelling at a WASA mast

• Mast data• OWC• Terrain around mast• GWC• WAsP modelling• Google Earth



Microscale modelling at WASA mast WM5

3 November 2016

• 10x10 km area• 100 m grid

resolution• AEP of a 2MW wind

turbine • Range: 5.5 to 10.5

GWh/year• Modelling quality of

the variability of resources is vital for wind energy projects and planning



Microscale modelling at all WASA masts

3 November 2016

WAsP resource grids from Observational Wind Atlas• 10 x 10 km grid• 100 meter grid spacing • Downloadable from project website

Wind speed ms-1 at 80 m above ground level



Microscale modelling on a “larger scale”

3 November 2016

The possibility to model at microscale around each of the masts gives us a very good tool for looking at the local wind resource.

So, isn’t a shame we don’t have these wind measurement masts all over South Africa?

But…wait a minute…that’s exactly what we do have now with the Verified Numerical Wind Atlas (at least for the project’s domain)

• Each one of the VNWA grid points represents a virtual wind measurement mast

• Each one has an associated Generalised Wind Climate data file

• Just like each of the 10 real masts



5 km

5 km

Example domain:• Terrain digitised

orography and roughness

• 9 “mesoscale” tiles• Each tile 5 x 5km




3 November 2016

If we combine the data from: With the technique from:

Then we can obtain a microscale-resolution resource grid over a larger area, further away from a real met mast

Generalized wind speed, h=100 m, z0=0.03 m



5 km

5 km

Example:• Actual met. mast• From measured data:

Observed Wind Climate

• From modelled terrain: Observed Wind Atlas




5 km

5 km

Example modelling:• Wind resource

predicted from mast-derived GWC




5 km

5 km

Example application:• Wind farm near mast

location• Predicted AEP • Predicted site

conditions according to IEC standards



5 km

5 km

Example NWA data:• Mesoscale modelling• 5 x 5 km grid• A NWA data ‘point’

within each grid cell• Each cell has an

associated NWA GWC




5 km

5 km

Example verification:• Compare GWC of the

Numerical Wind Atlas for the cell nearest to the GWC derived from the real met. mast measurements




5 km

5 km

Example “larger scale” mesoscale modelling:• Use the GWC from

each Numerical Wind Atlas cell

• Microscale model each tile or cell

• To obtain a 15 x 15km microscale wind resource map




5 km

5 km

Planning anywhere in WASA domain




5 km

5 km

• AEP of wind farm projects away from actual met mast measurements –bankability ??

• Comparison of wind data uncertainties

• Uncertainties depend on location and terrain type




Microscale vs. mesoscale: effect of resolution


KAMM/WAsP wind resource mapWind farm of 5 x 2 MW turbines

Grid cell size Estimated AEP

20 m 55 GWh

5120 m 39 GWh


Visualising micro and mesoscale

3 November 2016

• Site WM05• Created from Observational Wind Climate• Single dataset• Single colour grading




WAsP workspace example




Microscale modelling resource tiles overlaid onto a Google Earth image

Summary:• Input: WASA wind atlas

climatology• Elevation (user defined)• Roughness (user defined)• Any size and resolution




3 November 2016

We have over 40000 virtual masts in the WRF VNWA for South Africa but we cannot possibly microscale model all these together…

…well, actually we can.

• 40 000 GWC data from WASA

• Terrain and roughness mapping of whole WASA domain

• Google Earth

huge



Large scale wind resource map



Large scale wind resource map details

• Wind data from the Verified Numerical Wind Atlas– WRF mesoscale model, 3 km

• Elevation (for making contour maps)– 100-m elevation grids from Space Shuttle Topography Mission, SRTM

version 3.

• Land cover and roughness– USGS Global Land Cover Characteristics database.– Transformation table for z0

• Output is mean wind speed at 100m, resolution is 250m• As the modelling is done by WAsP, other outputs are also available e.g.

power density, terrain ruggedness index, etc. • Has also been verified using measured data from the 10 mast sites



Large scale power density map



Large scale terrain ruggedness index map



Verification using measurements @ 10 masts

• Comparison of actual wind speed predictions• Annual mean wind speeds at 100m a.g.l.• Site around masts modelled at microscale

3 November 2016

The OWCs at the mast sites were used to predict the wind speeds at

100m at the nearest calculation point for the WAsP Wind Resource Mapping Tool (“Frogfoot”). These were then compared to the large scale wind resource map values.

On average, the large scale wind resource map underestimates the

wind speed by just 1%



WASA project wind forecasting, etc.

3 November 2016

There’s one more product of the WASA project we should look at:

• Daily wind forecasts to understand wind regimes. WRF wind forecasts: http://veaonline.risoe.dk/wasa

• Setup and run of mesoscale models

• Verification against data from the 10 WASA met stations



G

Global Local

Global wind resources

Regional

Mesoscale modellingKAMM/WAsP, MM5, WRF, etc.

Microscale modelling(WAsP, other linear/nonlinear models)

200km x 200km 5km x 5km ~ 1-10 m

Summary of numerical wind atlas steps

Downscaling from global reanalysis data, verification and use



Summary of outputs from the WASA project

• Two state-of-the-art verified numerical wind atlases• High quality wind measurements from 10 met. masts• Associated observed and generalised wind climates => observational

wind atlas (and WAsP .wwh files)• A state-of-the-art extreme wind atlas• Large scale wind resource maps• Wind and wind power forecasting website

3 November 2016

Numerical wind atlases

Mesoscale modelling





Extreme wind atlas

Mesoscale modelling


Veri

ficat

ionLarge scale

wind resource maps

Forecasting web site



Course outline

3 November 2016




Accessing and using the WASA data

There are two main project websites for data and documents:• www.wasa.csir.co.za mainly for data, downloads and guidelines• www.wasaproject.info mostly news and information about the project

There is also a website for wind forecast data:• veaonline.risoe.dk/wasa/

WP4 Application Training Course 3 November 2016128


Main site for data and guidelines

WP4 Application Training Course

Main links:

• Information about each of the WASA met. masts and on-line data

• Map of mast locations

• Data download site

• Guide for accessing information

www.wasa.csir.co.za

3 November 2016129


WASA online data


www.wasa.csir.co.za

3 November 2016130


WASA online data


www.wasa.csir.co.za

3 November 2016131


Main site for download of WASA data


To download data either:

• Login

or

• Register if first-time user

• Access is free

Other links:

• Mast specifications and location reports

wasadata.csir.co.za/wasa1/WASAData

3 November 2016132


Downloading the WASA data


Downloadable data:

• Phase 1 final reports

• Wind time-series

• Extreme wind atlas

• Large scale wind resource maps

• Wind atlas data

• Data from met. masts

wasadata.csir.co.za/wasa1/WASAData

3 November 2016133




There are reports for:

• Met. stations and sites

• Measurements and data

• Observational wind atlas

• Numerical wind atlas

• Extreme wind atlas

• Planning & development of wind farms

• Guides to using the data

• Guide to applications

wasadata.csir.co.za/wasa1/WASAData > Final reports April 2014

3 November 2016134




Download of time-series:

• WRF simulations on 9×9 km2 grid.

• Data resolution 27 km

• Hourly time interval

• Period from 1990-2012

• Lambert conf. projection

wasadata.csir.co.za/wasa1/WASAData > Wind time-series downloads

3 November 2016135




Download of grids:

• 50-y wind speed

• 50-y gust speed

• Data resolution 0.05 deg

• ArcGIS ENVI format

• Datum: 94 Hartbeeshoek

wasadata.csir.co.za/wasa1/WASAData > Extreme wind atlas downloads

3 November 2016136




Download of time-series:

• WRF/WAsP simulations• Wind speed• Power density• Elevation• Ruggedness index

• Data resolution 250 m

• Average values only

• UTM or geographical

wasadata.csir.co.za/wasa1/WASAData > Large-scale wind resource map

3 November 2016137




Wind atlas downloads:

• Numerical wind atlas

• Observational wind atlas

• Case studies in SA

• WAsP workspaces

• Reports and guidelines

• Links, map data, tools

• Links to software

wasadata.csir.co.za/wasa1/WASAData > Wind atlas downloads

3 November 2016138




Measurements download:

• WASA Phase 1 stations: WM01-WM10 (2010- )

• WASA Phase 2 stations: WM11-WM15 (2015- )

• 10-min averaging time

• Data has been QA’ed

• OWC’s available too

wasadata.csir.co.za/wasa1/WASAData >Meteorological data downloads

3 November 2016139


News and information on the WASA project


Information about:

• Events

• WASA project• The WASA team

• Scope of the project and Work Packages

• Presentations of the project and results at workshops and seminars

www.wasaproject.info/index.html

3 November 2016140


WASA forecasting data


veaonline.risoe.dk/wasa/

3 November 2016141


Accessing and using the WASA data



Course outline

3 November 2016




Case Studies

Set of case studies designed to:

• Demonstrate specific uses of the WASA data• Show certain features of wind resource assessment• Indicate understanding needed when interpreting results• Be followed by course participants• Serve as inspiration for exploring more about wind atlases

The site(s) have been chosen for the above purpose. Please bear the following in mind:

• They are not necessarily a suggestion of a good wind farm location• Aspects other than wind resource have not been reviewed• Case studies are not intended to be a course in the use of wind resource

assessment software



Case study area: Overberg District Municipality



Case study 1: Planning

Planning for wind energy by a municipality

This case study demonstrates a typical use of the WASA data for planning and as input to assess the following questions:

Where are the good wind resources in the planning area so that we can:• Understand the potential for wind power - identify where the best areas

are physically located• Know more as background for choice of planning approach and

selection of zones for development• Investigate any possible relations to other land uses



Case study 1: slides

Planning for wind energy by a municipality• General wind resource in the area: NWA map• 100m height, uniform surface and uniform roughness




Planning for wind energy by a municipalityIdentifying LIB files




Planning for wind energy by a municipalityGeneral wind resource in the area: NWA map




Planning for wind energy by a municipality

More detailed wind resource in a particular area: microscale modelling using NWA

Steps:• Identify NWA nearest grid point and download LIB-files• Make terrain maps for the area Elevation (orography) Land cover (roughness)

• Import maps and wind data into modelling software (e.g. WAsP)• Run model to obtain resource maps as required




Planning for wind energy by a municipalityMore detailed wind resource in a particular area: microscale modelling using NWA



Case study 2: Wind farm energy yield

Wind farm development by a potential wind farm owner

This case study reviews estimates of the potential Annual Energy Production (AEP) of a specific wind farm and is designed to show how WASA data can help with input to answer among others the following question:

How do we start assessing the wind resource so we can make a good wind farm project?




Wind farm development by a potential wind farm owner (continued)

…or in other words:• How do we find the good resource areas?• Can candidate sites be identified? • How many turbines and what capacity?• What is a reasonable estimate of AEP?• Further stepsHow do we plan a wind measurement campaign to reduce uncertainty

and make the project bankable?How do we go about assessing the site conditions (turbulence intensity,

flow angles, wind shear, 50-year wind, etc) – design wind conditions, class of turbine?

• Other stepsHow do we get information about our possible wind farm so we can

begin to analyse the possibilities for grid connection, transport, planning permission, environmental impact , etc?




Potential good site identified from looking at wind resource map




Wind farm AEP according to WM05 measurements and NWAs based on WRF:

3 November 2016

input to WAsP Net AEP* “Stanford case”

LIB-file from WM05 measurements 149 GWh/y

Several LIB-files nearest to each WTG –WRF based NWA

117 GWh/y

Over-estimation when using WM05 27 %

* At simple assumptions incl. 100% availability, no losses, standard air density and standard WAsP parameters

OWA[m/s]

WRF[m/s]

DifferenceWRF

WM05 8.98 9.04 0.67%

NWA verifications at WM05: WARNING: Using the GWC derived from the WM05 mast significantly over-predicts the AEP. The similarity principle has not been adhered to: the mesoscale weather effects are not the same at the mast and the prediction site.



Case study 3: Extreme wind analysis


This case study briefly introduces the concept of site assessment that a developer needs to do in order to classify the site.

• Use data downloaded from the WASA Extreme Wind atlas to make an initial assessment of the site wind conditions.

• From this, the class of turbine suitable for the site can be determined from the IEC standard 61400-1

• Data for the mast location M5 can be downloaded and made into an generalised extreme wind climate (GEWC) format.

• Analysis software (e.g. WAsP Engineering) can be used to then model the extreme wind and turbulence intensity.





3 November 2016

Generalised Extreme Wind Climate imported into WEng

Site information concerning terrain is the same as that used in the WAsP modelling

Extreme 50 year return wind at 100m = 30.8m/s





3 November 2016

Turbulence intensity 0.119





3 November 2016

From the IEC 61400-1 table, the suitable class is IIIC



Case studies: concluding remarks

3 November 2016

Application for planning• NWA should be used directly as input to microscale modelling in order to

get sufficient detail and avoid the underestimation of wind energy potential due to mesoscale model averaging

• Manual approach or automated interpolation method for microscale modelling may be applied. In case of manual approach small tiles are recommended.

Application for wind farm studies• NWA LIB-files provide information about geographical variation of wind

resources that may have significant impact on resource assessment at areas with wind climate gradients – as shown in the case study

• It is recommended to consult the verification made for the NWA when assessing uncertainties in energy yield studies

• Does not negate the necessity for on-site measurements that will reduce uncertainties and thereby capital costs



Course outline

3 November 2016




WASA: Good practice use of data

All data have an uncertainty

But this doesn’t mean that they are not useful – just so long as we have a good idea of what that uncertainty is!

When considering the expected energy yield from a wind farm, some of the main areas of uncertainty are:• Wind measurement – instruments• Long term variation of wind• Flow modelling, especially wakes• Turbine power curve• Technical losses

For a land-based wind farm in terrain that is well understood then a total uncertainty estimate of between 7 and 15% would be reasonable.



Sources of uncertainty for energy yield

3 November 2016

H. Klug (2006). What does Exceedance Probabilities P90-P75-P50 Mean? DEWI Magazin Nr. 28, February 2006, 38-39.



Commonly used uncertainty factors



Uncertainty and the WASA data

3 November 2016

• Doing what a user would do: predicting the energy production

The graphic shows the predicted AEP from one 2MW wind turbine placed on the site of each of the

masts:• Prediction using NWA GWC• Prediction using OWC GWC

On average, the NWA over-predicts the energy production by around

6%, with a spread (standard deviation) of around 12%



Uncertainty and the WASA WRF NWA data

• Verification against measurements – This tells us something about the quality of the modelling– WRF-based NWA achieves a mean difference of around 4%– At the verification sites: a tendency to over-estimate of around 2%

• Uncertainty compared to measurement-based calculations – This concerns the application of the NWA data– Our application of the data gives around a 6% uncertainty on wind

speed prediction and about 12% uncertainty on energy prediction• Bankability – The confidence an investor has in a prediction of a

potential wind farm’s energy production– Still need dedicated site measurement (unless the site is close to a

mast but even here be sure the similarity principle holds)

The WASA project has reduced many uncertainties and removed others completely, especially by setting up the measurement masts and recording

wind data according to best practices.



Course outline

3 November 2016




Wrap-up and possible next steps

A wind atlas is:• A collection of generalised wind data• Generalised = flat terrain, uniform roughness, constant height a.g.l.• Independent of changes in terrain or land use

A numerical wind atlas is:• A wind atlas where the original wind data input is downscaled from

weather models

A verified numerical wind atlas is:• A wind atlas that has been compared to measured, high-quality data

A wind atlas can be used for:• Planning (e.g. zoning of wind farm areas, co-ordination with other

resources)• Reviewing production estimates from prospective wind farms• Planning wind farm projects



The WASA project products

3 November 2016

2 x VNWAs

Extreme wind atlas10 x met. masts and data

Forecasting

Microscale modelling and wind

resource maps



Where to find information

The two main project websites for data and documents:• http://www.wasa.csir.co.za mainly for data and guidelines• http://www.wasaproject.info/index.html mostly news and information

about the project

Website for forecast data:• http://veaonline.risoe.dk/wasa/



Possible next steps

Using WASA data• Rehearse case studies when you get back to your desk• Try using WASA data for your own purposes

Further information• Look at documentation on the websites

Further training:• WAsP training course (http://www.wasp.dk/)• Online WAsP course• WAsP Engineering course



The WASA Project Team

3 November 2016

SANEDI South African National Energy Development Institute

• executing agency – contracting the implementing partners• coordination and dissemination

UCT CSAG Climate System Analysis Group, University of Cape Town

• mesoscale modelling

CSIR Built Environment, Council for Scientific and Industrial Research

• measurements and microscale modelling

SAWS South African Weather Service

• extreme wind assessment

DTU Wind Energy Dept of Wind Energy, Technical University of Denmark

• partner in all activities



The WASA Project Training Material Team

3 November 2016

DTU Wind Energy Dept of Wind Energy, Technical University of Denmark

• Tom Cronin, Niels G. Mortensen, Andrea Hahmann, Jens Carsten Hansen, Jake Badger, Morten Nielsen & Xiaoli Larsen

UCT CSAG Climate System Analysis Group, University of Cape Town

• Chris Lennard

CSIR Built Environment, Council for Scientific and Industrial Research

• Eugéne Mabille and Eric Prinsloo

SAWS South African Weather Service

• Andries Kruger





PlannersLocal authoritiesStudents

THE END


Documents

Teaching wind resources, wind energy planning and projects ...€¦ · 3 November 2016 This course has been put together for the purposes of teaching planners, assessors and students