remote sensing
Article
Wind Resource Assessment for High-Rise BIWT
Using RS-NWP-CFD
Hyun-Goo Kim 1, *, Wan-Ho Jeon 2 and Dong-Hyeok Kim 3
1
2
3
*
Korea Institute of Energy Research, Daejeon 34129, Korea
CEDIC Co. Ltd., Seoul 08506, Korea; [email protected]
Chungnam Institute, Gongju 32589, Korea; [email protected]
Correspondence: [email protected]; Tel.: +82-42-860-3376
Academic Editors: Charlotte Bay Hasager, Alfredo Peña, Xiaofeng Li and Prasad S. Thenkabail
Received: 29 July 2016; Accepted: 8 December 2016; Published: 13 December 2016
Abstract: In this paper, a new wind resource assessment procedure for building-integrated wind
turbines (BIWTs) is proposed. The objective is to integrate wind turbines at a 555 m high-rise
building to be constructed at the center of Seoul, Korea. Wind resource assessment at a high
altitude was performed using ground-based remote sensing (RS); numerical weather prediction
(NWP) modeling that includes an urban canopy model was evaluated using the remote sensing
measurements. Given the high correlation between the model and the measurements, we use the
model to produce a long-term wind climate by correlating the model results with the measurements
for the short period of the campaign. The wind flow over the high-rise building was simulated using
computational fluid dynamics (CFD). The wind resource in Seoul—one of the metropolitan cities
located inland and populated by a large number of skyscrapers—was very poor, which results in a
wind turbine capacity factor of only 7%. A new standard procedure combining RS, NWP, and CFD is
proposed for feasibility studies on high-rise BIWTs in the future.
Keywords: building-integrated wind turbine; wind resource assessment; numerical weather
prediction; computational fluid dynamics; Seoul
1. Introduction
In the Bible, the Tower of Babel was a symbol of the aspiration of humans to reach God’s realm.
Pyramids built by the ancient Egyptians were also a reflection of such desire. The Great Pyramid
of Giza, whose height is 145 m and which was built around BC 2560, was the tallest building in the
world until it was replaced by the 160 m high Lincoln Cathedral in England in AD 1311. Since then the
emergence of the highest buildings has been chronicled since then as follows: Eiffel Tower in Paris,
France (300 m) in 1889; Empire State Building in New York, NY, USA (381 m) in 1930; Ostankino Tower
in Moscow, Russia (537 m) in 1967; CN Tower in Toronto, Canada (553 m) in 1975; and Burj Khalifa in
Dubai, UAE (830 m) in 2007 [1].
This landmark race continues in southeast and east Asia as well: Petronas Twin Towers in Kuala
Lumpur, Malaysia (452 m) in 1996; Jin Mao Tower in Shanghai, China (421 m) in 1998; Taipei World
Financial Center 101 in Taipei, Taiwan (509 m) in 2005; and Lotte World Tower in Seoul, Korea (555 m)
in 2016.
Note that the concept of sustainable architecture, which is contrary to ultra high-rise architectures,
started in the 21st century. Sustainable architecture refers to the conscious minimization of the
negative effects on the environment by making building materials, energy use, and space utilization
efficient so that the actions and decisions we make today will not limit the opportunities for the
next generations [2,3]. In other words, a new paradigm wherein buildings shall provide symbolism
such as ecological design or sustainability—instead of concentrating on the tallest building race—has
Remote Sens. 2016, 8, 1019; doi:10.3390/rs8121019
www.mdpi.com/journal/remotesensing
Remote Sens. 2016, 8, 1019
2 of 17
tallest building race—has begun. As a result, the so-called building-integrated ideas, wherein
begun.
As aenergy
result,as
the
so-called
ideas, wherein
renewable
energy
as a symbol
of
renewable
a symbol
ofbuilding-integrated
sustainable and eco-friendly
energy
is integrated
in buildings,
have
sustainable
and
eco-friendly
energy
is
integrated
in
buildings,
have
been
implemented.
been implemented.
The
World Trade
The Bahrain
Bahrain World
Trade Center,
Center, which
which was
was completed
completed in
in 2008
2008 and
and is
is regarded
regarded as
as the
the first
first
modern
BIWT,
is
a
240
m-high,
50-story
twin
tower
complex.
It
is
a
structure
with
symmetrical
modern BIWT, is a 240 m-high, 50-story twin tower complex. It is a structure with symmetrical
triangular-shaped
shape
andand
layout
werewere
designed
to make
use of use
windofenergy
triangular-shaped twin
twinbuildings
buildingswhose
whose
shape
layout
designed
to make
wind
beyond
mere integration
of wind turbine
in the building.
It isthe
regarded
as building-augmented
energy the
beyond
the mere integration
of wind
turbine in
building.
It is regarded as
wind
turbine (BAWT),wind
which
is a more
aggressive
than
BIWT [4].concept than BIWT [4].
building-augmented
turbine
(BAWT),
whichconcept
is a more
aggressive
As
result via
via CFD
CFD (by
(by the
the authors;
authors; see
see Figure
Figure 1),
1),
As shown
shown in
in the
the building
building photo
photo and
and on
on analysis
analysis result
3-blade
horizontal-axis
wind
turbines
with
diameter
of
29
m
and
capacity
of
225
kW
were
installed
3-blade horizontal-axis wind turbines with diameter of 29 m and capacity of 225 kW were installed
at
the bridges
bridges connecting
connectingthe
thetriangular
triangulartwin
twintowers,
towers,which
whichare
aresymmetrical
symmetrical
with
interior
angle
at the
with
anan
interior
angle
of
◦
of
about
120
.
Since
the
building
is
located
at
the
coast,
the
Venturi
effect
occurs
wherein
in-between
about 120°. Since the building is located at the coast, the Venturi effect occurs wherein in-between
distance
reduced by
byan
anangle
anglemade
madebybythe
the
twin
towers
when
breeze
blows.
building
distance reduced
twin
towers
when
thethe
seasea
breeze
blows.
The The
building
was
was
designed
to
increase
the
efficiency
of
wind
power
generation
through
the
accelerated
wind.
designed to increase the efficiency of wind power generation through the accelerated wind. About
About
1.3per
GWh
per year—which
forthe
13%
of theconsumption
energy consumption
of the building—is
1.3 GWh
year—which
accountsaccounts
for 13% of
energy
of the building—is
generated
generated
by
wind
power,
for
capacity
factor
of
approximately
22%.
by wind power, for capacity factor of approximately 22%.
(a)
(b)
Figure 1. The Bahrain World Trade Center: (a) photo from the front of the wind turbines; and (b)
Figure 1. The Bahrain World Trade Center: (a) photo from the front of the wind turbines; and (b) contour
contour plot of wind speed on a horizontal cross-sectional plane at 100 m aboveground (CFD by the
plot of wind speed on a horizontal cross-sectional plane at 100 m aboveground (CFD by the authors).
authors).
The
As aa 147
147 m-high,
m-high,
The Hess
Hess Tower
Tower in
in Houston,
Houston, Texas,
Texas, USA
USA was
was completed
completed in
in 2009
2009 (Figure
(Figure 2).
2). As
20-story
building,
it
was
regarded
as
BIWT
wherein
10
vertical-axis
wind
turbines
were
installed
20-story building, it was regarded as BIWT wherein 10 vertical-axis wind turbines were installed on
on
top
drew attention
attention because
because aa fragment
fragment of
of the
the part
part of
top [5].
[5]. This
This building
building drew
of aa wind
wind turbine
turbine damaged
damaged by
by
strong
event
being
covered
in the
community
newspaper.
The Hess
strong wind
windfell,
fell,with
withthe
thedangerous
dangerous
event
being
covered
in local
the local
community
newspaper.
The
Tower
case
suggests
that
BIWTs
installed
in
high-rise
buildings
should
consider
all
accident
factors
Hess Tower case suggests that BIWTs installed in high-rise buildings should consider all accident
including
gusts and
lightning.
factors including
gusts
and lightning.
The
River
Tower
in Guangzhou,
China is aChina
typical example
of BIWTs
and building-integrated
The Pearl
Pearl
River
Tower
in Guangzhou,
is a typical
example
of BIWTs and
photovoltaic
(BIPV)
in
Asia.
As
a
309
m-high,
71-story
building
completed
in
2011,
it was
designed to
building-integrated photovoltaic (BIPV) in Asia. As a 309 m-high, 71-story building
completed
in
increase
the
wind
speed
via
the
nozzle
effect
by
making
flow
passages
in
the
building
as
shown
in
2011, it was designed to increase the wind speed via the nozzle effect by making flow passages in the
Figure
3. A
of four
8 m-high
vertical-axis
wind
turbinesvertical-axis
are installedwind
insideturbines
the building,
each at
building
astotal
shown
in Figure
3. A
total of four
8 m-high
are installed
the
internal
passage
on
the
left
and
right
sides
[6].
inside the building, each at the internal passage on the left and right sides [6].
Remote Sens. 2016, 8, 1019
3 of 17
Remote Sens. 2016, 8, 1019
Remote Sens. 2016, 8, 1019
3 of 17
3 of 17
(a)
(a)
(b)
(b)
Figure 2. The Hess Tower: (a) photo from the front of the wind turbines; and (b) contour plot of wind
Figure 2. The Hess Tower: (a) photo from the front of the wind turbines; and (b) contour plot of wind
Figureon
2. aThe
Hess cross-sectional
Tower: (a) photo
from(CFD
the front
of authors).
the wind turbines; and (b) contour plot of wind
speed
vertical
plane
by the
speed on a vertical cross-sectional plane (CFD by the authors).
speed on a vertical cross-sectional plane (CFD by the authors).
(a)
(a)
(b)
(b)
(c)
(c)
Figure 3. The Pearl River Tower: (a) photo of front view; (b) wind penetration structure; and (c)
Figure 3.
The Pearl
River
Tower:
(a) photo of front view;
(b) wind
structure; and (c)
contour
wind
speed
on a (a)
horizontal
plane
(CFD penetration
by thestructure;
authors).
Figure
3.plot
Theof
Pearl
River
Tower:
photo ofcross-sectional
front view; (b) wind
penetration
and (c) contour
contour plot of wind speed on a horizontal cross-sectional plane (CFD by the authors).
plot of wind speed on a horizontal cross-sectional plane (CFD by the authors).
Built in 2011, the Strata SE1 in London is a 148 m-high, 43-story residential building wherein
in 2011,
the Strata SE1
London with
is a 148
m-high,
43-story
wherein
threeBuilt
rooftop
horizontal-axis
windinturbines
19 kW
capacity
wereresidential
integrated building
(Figure 4).
It has
Built
in
2011,
the
Strata
SE1
in
London
is
a
148
m-high,
43-story
residential
building
wherein
threeawarded
rooftop horizontal-axis
wind design
turbines
with but
19 kW
capacity
integrated claiming
(Figure 4).
Itthree
has
been
several architectural
prizes,
the press
waswere
not impressed,
that
the
rooftop
horizontal-axis
wind
turbines
with
19
kW
capacity
were
integrated
(Figure
4).
It
has
been
been
awarded
several
architectural
design
prizes,
but
the
press
was
not
impressed,
claiming
that
the
wind turbines rarely rotated. The citizens believe the wind turbine should work continuously, but
awarded
several
architectural
design
prizes,believe
but thethe
press
wasturbine
not impressed,
claiming
that the wind
wind
turbines
rarely
rotated.
The
citizens
wind
should
work
continuously,
but
the wind is generally low [7].
turbines
rotated.low
The[7].
citizens believe the wind turbine should work continuously, but the wind
the windrarely
is generally
is generally low [7].
Remote Sens. 2016, 8, 1019
4 of 17
Remote Sens. 2016, 8, 1019
4 of 17
(a)
(b)
Figure 4. The Strata SE1 Apartment: (a) photo from the rear of the wind turbines; and (b) contour
Figure 4. The Strata SE1 Apartment: (a) photo from the rear of the wind turbines; and (b) contour
plot of wind speed on a horizontal cross-sectional plane (top) and streamlines passing through holes
plot of wind speed on a horizontal cross-sectional plane (top) and streamlines passing through holes
(bottom; CFD by the authors).
(bottom; CFD by the authors).
This study was conducted to integrate renewable energy in the Lotte World Tower, a 555 m
This building
study was
conducted
to integrate
renewable
in thebuilding
Lotte World
555 m
high-rise
to be
constructed
in Seoul, Korea
as theenergy
sixth tallest
in theTower,
world.aSeoul
is
high-rise
building
to
be
constructed
in
Seoul,
Korea
as
the
sixth
tallest
building
in
the
world.
Seoul
is
one of the largest metropolitan cities with around 10 million citizens. Seoul is characterized by its
one
of the
largest
metropolitan
citieswind
with compared
around 10 to
million
citizens. Seoul is characterized by its
inland
location
as well
as weakening
onshore.
inland
location
as
well
as
weakening
wind
compared
to
onshore.
In this study, wind resource assessment was performed to evaluate the performance of a 555 m
In
this
study,
resource
was performed
to evaluate
the complicated
performanceto
of perform
a 555 m
high-rise
BIWT
to wind
be built
at the assessment
center of Seoul.
At such height,
it is very
high-rise
BIWT to be built
at the center
of Seoul. At remote
such height,
it is such
very complicated
to perform
in-situ measurements.
Therefore,
ground-based
sensors
as LIDAR and
SODARin-situ
were
measurements.
Therefore,
ground-based
remote
sensors
such
as
LIDAR
and
SODAR
were
employed
employed to perform wind resource assessment. Furthermore, a numerical weather prediction
to
perform
wind
resource
assessment. Furthermore,
a numerical
(NWP)
model
(NWP)
model
and
measure-correlate-predict
(MCP) method
wereweather
utilizedprediction
to extend the
short-term
and
measure-correlate-predict
(MCP)
method
were
utilized
to
extend
the
short-term
data
measured
data measured via remote sensing to more than one-year data. A three-dimensional geographical
via
remote sensing
moredatabase
than one-year
data.
A three-dimensional
information
system
information
systemto(GIS)
and CFD
were
also employed to geographical
predict the effect
of the building
(GIS)
on thedatabase
flow. and CFD were also employed to predict the effect of the building on the flow.
2. Analysis Methods
2. Analysis Methods
2.1. Analysis Procedure
2.1. Analysis Procedure
Figure 5 shows the procedure for wind resource assessment to implement BIWTs at the upper
Figure
5 shows
the procedure for wind resource assessment to implement BIWTs at the upper
part of
high-rise
buildings.
part of
high-rise
The
remotebuildings.
sensing campaign in this study lasted for two months only due to various
The
remote
sensing
campaign
in this
studyIn lasted
for an
twoNWP,
months
onlyWeather
due toResearch
various
constraints.
An MCP
method
is therefore
applied.
this study,
namely
constraints. An MCP
method
is therefore
applied.
In used
this study,
an NWP,against
namelythe
Weather
Research
Forecasting–Urban
Canopy
Model
(WRF–UCM)
was
and evaluated
RS data.
The RS
Forecasting–Urban
Canopy
Model
(WRF–UCM)
was
used
and
evaluated
against
the
RS
data.
The
data were then utilized as reference data, and long-term wind climate data were reconstructed through
RS
data
were
then
utilized
as
reference
data,
and
long-term
wind
climate
data
were
reconstructed
an MCP method.
through
MCPstep
method.
The an
second
is to create a wind resource map that can show the spatial variability of the wind
The
second
step
is to area
create
wind resource
map that
can show the spatial variability of the
resource at the installation
to adetermine
the turbines’
layout.
windFor
resource
at
the
installation
area
to
determine
the
turbines’
wind resource mapping, first, wind flow fields at thelayout.
target area by wind direction are
For wind
mapping,
first, wind
flow are
fields
at the target
area
wind directions
direction are
are
simulated
usingresource
CFD. If 16
wind directional
sectors
considered,
a total
of by
16 wind
simulated
using
CFD.
If
16
wind
directional
sectors
are
considered,
a
total
of
16
wind
directions
are
to be simulated. Next, 16 simulation results are synthesized by applying frequencies of occurrence
to be
simulated.
Next,
16 simulation
are synthesized
by applying
of occurrence
by
each
of the wind
directions
and 1 results
m/s wind
speed intervals
obtained frequencies
from the wind
climate as
by
each
of
the
wind
directions
and
1
m/s
wind
speed
intervals
obtained
from
the
wind
climate as
weighting factors. The procedure can be described as:
weighting factors. The procedure can be described as:
16 25
V ( x, y, z) = ∑ ∑ Vij ( x, y, z)· f ij
, ,
, , ∙
i =1 j =1
(1)
(1)
Remote Sens. 2016, 8, 1019
Remote Sens. 2016, 8, 1019
5 of 17
5 of 17
Remote Sens. 2016, 8, 1019
5 of 17
where,
V ( x, y,, z,) is is
the
synthesized
windatataadirectional
directionalsector
sector
i and
is the
the wind
i and
where,
the
synthesizedwind
wind speed,
speed, Vij ( x,, y,, z) is
wind
speed
bin
j,
and
f
is
the
frequency
of
occurrence
by
each
directional
and
wind
speed
bin.
ij synthesized
wind
speed ,bin
is the frequency
of occurrence
directional
wind speed
bin.
, by
, each
is the
wind at aand
directional
sector
i and
where,
, j, and
is the
wind speed,
The
third
step
is
to
select
a
type
of
wind
turbine
and
determine
an
optimum
layout.
third
is to select
type of wind
turbine and
an optimum
layout.
Although
windThe
speed
binstep
j, and
is thea frequency
of occurrence
bydetermine
each directional
and wind
speed
bin.
Although
energy
production
can
be
maximized
if
a
wind
turbine
is
installed
at
the
highest
wind
energy
can
maximized
a wind
turbine
is installedanatoptimum
the highest
wind
power
Theproduction
third step is
to be
select
a type of if
wind
turbine
and determine
layout.
Although
power
density
location,
wake
losses
shall
also
be
taken
into
account.
density
wake
shall also beif taken
intoturbine
account.
energy location,
production
canlosses
be maximized
a wind
is installed at the highest wind power
density location, wake losses shall also be taken into account.
Figure 5. Procedure of wind resource assessment for a high-rise BIWT by using RS–NWP–CFD.
Figure 5. Procedure of wind resource assessment for a high-rise BIWT by using RS–NWP–CFD.
Figure 5. Procedure of wind resource assessment for a high-rise BIWT by using RS–NWP–CFD.
2.2. Remote Sensing Campaign
2.2. Remote Sensing Campaign
2.2. Remote
Sensingsensing
Campaign
The remote
campaign, which consisted on measuring the wind resource at 555 m
The
remote
sensing
campaign,
which(yellow
consisted
on measuring
the wind
resource
at 555 m270
altitude,
altitude,
was
performed
at
the rooftop
triangleon
in measuring
Figure 6) ofthe
Lotte
Hotel,
whichatis 555
m
The remote sensing campaign,
which consisted
wind
resource
was
performed
at
the
rooftop
(yellow
triangle
in
Figure
6)
of
Lotte
Hotel,
which
is
270
m
away
from
away
from
the
lot
(dashed
line
in
Figure
6)
of
the
Lotte
World
Tower
in
the
west
direction
near
the
altitude, was performed at the rooftop (yellow triangle in Figure 6) of Lotte Hotel, which is 270 m
theHan
lot (dashed
linecenter
in Figure
6) ofAthe
Lotte World
TowerLIDAR
in the west
direction
near MPAS
the Han
River at
the
of Seoul.
Leosphere
WLS70
Sintec
SODAR
awayRiver
fromatthe
lot (dashed
line in
Figure
6) ofWindCube
the Lotte World
Tower
inand
the awest
direction near
the
thewere
center
of
Seoul.
A
Leosphere
WindCube
LIDAR
WLS70
and
a
Sintec
MPAS
SODAR
were
deployed
deployed
the rooftop
of the
145 m-high
Lotte Hotel,
and WLS70
measurements
wereMPAS
conducted
for
Han River
at theatcenter
of Seoul.
A Leosphere
WindCube
LIDAR
and a Sintec
SODAR
at two
the rooftop
of at
the
145
m-high
Lotte
measurements
were conducted
two months
months
from
9the
March
to 8ofMay
2009.
were
deployed
rooftop
the
145Hotel,
m-highand
Lotte
Hotel, and measurements
were for
conducted
for
from
9
March
to
8
May
2009.
two months from 9 March to 8 May 2009.
Figure 6. Virtual landscape around the Lotte World Tower in Seoul (white dashed line indicates the
Lotte
Tower
under construction,
and yellow
indicates
the RSdashed
campaign
FigureWorld
6. Virtual
landscape
around the Lotte
Worldtriangle
Tower in
Seoul (white
linesite).
indicates the
Figure 6. Virtual landscape around the Lotte World Tower in Seoul (white dashed line indicates the
Lotte World Tower under construction, and yellow triangle indicates the RS campaign site).
World Weather
Tower under
construction, and yellow triangle indicates the RS campaign site).
2.3.Lotte
Numerical
Prediction
2.3. Numerical
Weather
Prediction
Since wind
resource
assessment requires measurements of at least one year including seasonal
2.3. Numerical Weather Prediction
variability,
short-term
remote
sensingrequires
measurements
are extended
through
an including
MCP method
that
Since wind resource assessment
measurements
of at least
one year
seasonal
Since
wind
resource
assessment
requires
measurements
of
at
least
one
year
including
seasonal
uses
NWP data.
NWP was
conducted
the urban
canopy layer
in order
to secure
reliable
variability,
short-term
remote
sensingconsidering
measurements
are extended
through
an MCP
method
that
variability,
sensing considering
measurements
are extended
through
an MCP
method
that
reference
data.
uses NWPshort-term
data. NWPremote
was conducted
the urban
canopy layer
in order
to secure
reliable
uses
NWP data.
reference
data. NWP was conducted considering the urban canopy layer in order to secure reliable
reference data.
Remote Sens. 2016, 8, 1019
Remote Sens. 2016, 8, 1019
6 of 17
6 of 17
WRF–UCM is
that
takes
into
account
thethe
interaction
of the
is aamesocale-microscale
mesocale-microscalecoupled
coupledmodel
model
that
takes
into
account
interaction
of
physical
processes
in aninurban
canopy
[8]. The
assumes
two-dimensional
buildings
and road
the physical
processes
an urban
canopy
[8]. model
The model
assumes
two-dimensional
buildings
and
types
for detailed
city landuse.
The single-layer
land surface
the lower
boundary
condition
road types
for detailed
city landuse.
The single-layer
land model,
surfacei.e.
model,
i.e. the
lower boundary
of
WRF–UCM,
employed the
Seoul building
height
layerheight
provided
byprovided
the National
Spatial
Information
condition
of WRF–UCM,
employed
the Seoul
building
layer
by the
National
Spatial
Clearinghouse
of
Korea.
Here,
the
lower
boundary
condition
was
modified
by
applying
the
surface
Information Clearinghouse of Korea. Here, the lower boundary condition was modified by applying
roughness
proposed
byproposed
Macdonald
al. (1998) [9].
Specifically,
displacement
the surfaceestimation
roughnessmethod
estimation
method
byetMacdonald
et al.
(1998) [9].the
Specifically,
the
height
(do ) of the
urban
calculated
to theand
terrain
elevation.
In addition,
surface
displacement
height
(docanopy
) of the was
urban
canopy and
was added
calculated
added
to the terrain
elevation.
In
roughness
(zo ) was
calculated;
it was
larger than
the
roughness
height
at the
landdefined
cover, the
addition, surface
roughness
(zoif) was
calculated;
if it
was
larger than
thedefined
roughness
height
at
value
wascover,
substituted.
the land
the value was substituted.
2.4.
Computational Fluid
Fluid Dynamics
2.4. Computational
Dynamics
The
around
thethe
Lotte
World
Tower
will bewill
considerably
distorted
once the once
building
The wind
windflow
flowfield
field
around
Lotte
World
Tower
be considerably
distorted
the
is
complete.
The
wind
flow
field
was
simulated
using
the
Reynolds-Averaged
Navier-Stokes
(RANS)
building is complete. The wind flow field was simulated using the Reynolds-Averaged
CFD
software SC/Tetra
a flow solver
on aisfinite
volume
method
Navier-Stokes
(RANS) v12.
CFDSC/Tetra
softwareisSC/Tetra
v12. based
SC/Tetra
a flow
solver
basedand
onemploys
a finite
the
SIMPLEC
algorithm
for pressure-velocity
coupling
and the
QUICK upwind
scheme
for
volume
method
and employs
the SIMPLEC
algorithm
for 2nd-order
pressure-velocity
coupling
and the
convection
terms.
2nd-order QUICK upwind scheme for convection terms.
The
draft of
of the
the Lotte
Lotte World
World Tower
Tower is
shape with
with square
square cross
cross section,
section,
The design
design draft
is aa tapered
tapered column
column shape
whose
edges
are
round
as
shown
in
Figure
7.
The
cross-sectional
area
decreases
in
a
stepwise
manner
whose edges are round as shown in Figure 7. The cross-sectional area decreases in a stepwise
with
increasing
altitude. This
designThis
is intended
to prevent
by shifting
the by
Karman
vortex
manner
with increasing
altitude.
design is
intendedresonance
to prevent
resonance
shifting
the
shedding
frequency
by altitude.
Theby
rooftop,
where
the windwhere
turbines
supposed
besupposed
installed,
Karman vortex
shedding
frequency
altitude.
The rooftop,
the are
wind
turbinestoare
has
a truss
structure
of a square
cone at of
theacenter
16 columns
surrounding
outer
to be
installed,
has consisting
a truss structure
consisting
squareand
cone
at the center
and 16 the
columns
border.
From
the
top
view,
the
lower
structures
protruding
on
three
sides
excluding
the
left
side
are
surrounding the outer border. From the top view, the lower structures protruding on three sides
shelves
to protect
by preventing
downwash
fromby
flowing
down downwash
the buildingfrom
as a result
of
excluding
the leftpedestrians
side are shelves
to protect
pedestrians
preventing
flowing
wind
colliding
with
the
building.
down the building as a result of wind colliding with the building.
Figure 7.
7. Dimensions
design and
and rooftop
rooftop structure:
structure: (left)
Figure
Dimensions of
of the
the Lotte
Lotte World
World Tower
Tower design
(left) side
side view;
view;
(right-top)
perspective
view
of
rooftop
structure;
and
(right-bottom)
top
view.
(right-top) perspective view of rooftop structure; and (right-bottom) top view.
The computational domain was set to 6 km × 6 km × 2.5 km around the Lotte World Tower and
The computational domain was set to 6 km × 6 km × 2.5 km around the Lotte World Tower and
Figure 8 shows the central part of the domain (3 km × 1.5 km area). The 90 million unstructured
Figure 8 shows the central part of the domain (3 km × 1.5 km area). The 90 million unstructured hybrid
hybrid mesh is used to model complex urban buildings. For a validation simulation, the main wind
mesh is used to model complex urban buildings. For a validation simulation, the main wind direction
direction (WNW) with/without the Lotte World Tower was simulated. Considering the periodic
(WNW) with/without the Lotte World Tower was simulated. Considering the periodic geometrical
geometrical characteristics of the urban canopy and a relatively short computational domain,
characteristics of the urban canopy and a relatively short computational domain, periodic boundary
periodic boundary conditions were imposed on inlet and outlet planes instead of an atmospheric
conditions were imposed on inlet and outlet planes instead of an atmospheric boundary layer inlet
boundary layer inlet condition. The constant pressure condition is imposed on the upper free
condition. The constant pressure condition is imposed on the upper free boundary and slip conditions
boundary and slip conditions are assigned to side boundaries. The WRF–UCM result was used as an
are assigned to side boundaries. The WRF–UCM result was used as an initial condition.
initial condition.
The CFD analysis employed the modified production k–ε turbulence model, which predicts flow
The CFD analysis employed the modified production k–ɛ turbulence model, which predicts
separation in the cubic block structures as shown in urban streets [10,11].
flow separation in the cubic block structures as shown in urban streets [10,11].
Remote
2016, 8,
Remote Sens.
Sens. 2016,
8, 1019
1019
Remote Sens. 2016, 8, 1019
of 17
17
77 of
7 of 17
Figure 8. Computation domain around the Lotte World Tower (left); and the pedestrian-level wind
Figure
8. Computation
flow
streamlines
(right).domain around the Lotte World Tower (left); and the pedestrian-level wind
Figure 8. Computation domain around the Lotte World Tower (left); and the pedestrian-level wind
flow streamlines (right).
flow streamlines (right).
3. Analysis Results
3. Analysis Results
3. Analysis Results
3.1. Remote Sensing Campaign
3.1. Remote Sensing Campaign
3.1. Remote
FigureSensing
9 showsCampaign
the mean wind speed profiles measured using LIDAR (from 10-min data) and
Figure
9 shows
the mean
wind
speed profiles
measured
using
LIDAR
(from 10-min
data)
and
SODAR
(from
30-min
for
the campaign
period,
where using
the data
recovery
(DRR;
dashed
Figure 9 shows thedata)
mean
wind
speed profiles
measured
LIDAR
(fromrate
10-min
data)
and
SODAR
(from
30-min
data)
for
the
campaign
period,
where
the
data
recovery
rate
(DRR;
dashed
lines) decreased
with data)
increasing
Slight
distortion
of wind
shear over
m above
ground
SODAR
(from 30-min
for thealtitude.
campaign
period,
where the
data recovery
rate400
(DRR;
dashed
lines)
lines)isdecreased
with
increasing
altitude. the
Slight
distortion
of with
windwhite
shearcircles)
over 400
m the
above
ground
level
observed
in
LIDAR
data
between
raw
data
(lines
and
concurrent
decreased with increasing altitude. Slight distortion of wind shear over 400 m above ground level
level (lines
is observedblack
in LIDAR data
between
the rawof
data
with white
circles)
and the concurrent
data
which
is
about
the(lines
raw
SODAR
data,
difference
is
observedwith
in LIDAR circles)
data between
the
raw 50%
data (lines
withdata.
whiteIncircles)
and
theminor
concurrent
data
data
(lines
with
black
circles)
which
is
about
50%
of
the
raw
data.
In
SODAR
data,
minorisdifference
is
shown
between
the
raw
and
concurrent
data
below
500
m
aboveground
where
DRR
over 50%
(lines with black circles) which is about 50% of the raw data. In SODAR data, minor difference
is
is shown between the raw and concurrent data below 500 m aboveground where DRR is over 50%
[12].
shown
between the raw and concurrent data below 500 m aboveground where DRR is over 50% [12].
[12].
(a)
(a)
(b)
(b)
Figure 9. Mean wind speed and data recovery rate profiles observed by the remote sensing
Figure 9. Mean wind speed and data recovery rate profiles observed by the remote sensing equipment
Figure 9. Mean
and data
recovery
rate (10-min
profiles readings);
observed and
by the
remote (30-min
sensing
equipment
(meanwind
for thespeed
campaign
period):
(a) LIDAR
(b) SODAR
(mean for the campaign period): (a) LIDAR (10-min readings); and (b) SODAR (30-min readings).
equipment (mean for the campaign period): (a) LIDAR (10-min readings); and (b) SODAR (30-min
readings).
readings).
3.2. Numerical Weather Prediction
3.2. Numerical Weather Prediction
3.2. Numerical
Prediction
Figure 10 Weather
shows the
simulation results at 12:00 p.m. when the westward wind blew to the inland
Figure 10 shows the simulation results at 12:00 p.m. when the westward wind blew to the
fromFigure
the west
sea.
The
figure
shows a results
difference
in wind
speed
between
two cases,
either
applying
the simulation
12:00
p.m.
westward
blew
to
the
inland from 10
theshows
west sea.
The figure shows
aatdifference
in when
wind the
speed
betweenwind
two cases,
either
the
urban
canopy
model
or
not.
Wind
speed
in
Seoul—marked
by
the
dashed
white
circle—shows
inland from
westcanopy
sea. The
figureorshows
a difference
speed between
twodashed
cases, either
applying
thethe
urban
model
not. Wind
speed in
in wind
Seoul—marked
by the
white
clear
reduction.
applying
the
urban
canopy
model
or
not.
Wind
speed
in
Seoul—marked
by
the
dashed
white
circle—shows clear reduction.
Figure 11 shows
the comparison among mean wind speed profiles measured by the 1-h averaged
circle—shows
clear
reduction.
Figure 11 shows the comparison among mean wind speed profiles measured by the 1-h
LIDAR
and 11
SODAR
and
results
using
1-h wind
averaged
WRF
and WRF–UCM
during
Figure
shows
thesimulation
comparison
among
mean
profiles
measured
by
the the
1-h
averaged
LIDAR
and SODAR
and simulation
results
usingspeed
1-h averaged
WRF
and WRF–UCM
remote
sensing
campaign
period.
The
LIDAR
profiles
agree
better
with
the
WRF–UCM
results
than
averaged LIDAR and SODAR and simulation results using 1-h averaged WRF and WRF–UCM
the SODAR profiles, while WRF results are noticeably the highest.
Remote
Remote Sens.
Sens. 2016,
2016, 8,
8, 1019
1019
88 of
of 17
17
during
the
remote
sensing
during
the2016,
remote
sensing campaign
campaign period.
period. The
The LIDAR
LIDAR profiles
profiles agree
agree better
better with
with the
the WRF–UCM
WRF–UCM
Remote
Sens.
8, 1019
8 of 17
results
than
the
SODAR
profiles,
while
WRF
results
are
noticeably
the
highest.
results than the SODAR profiles, while WRF results are noticeably the highest.
10.Wind
Wind
speed
difference
between
WRFWRF–UCM
and WRF–UCM
simulation
at 12:00
p.m.
Figure
speed
difference
between
WRF
simulation
at
(negative
Figure 10.
10.
Wind
speed
difference
between
WRF and
and
WRF–UCM
simulation
at 12:00
12:00 p.m.
p.m.
(negative
(negative
value
means
that
WRF
over-predicted
wind
speed
than
WRF–UCM).
value
means
that
WRF
over-predicted
wind
speed
than
WRF–UCM).
value means that WRF over-predicted wind speed than WRF–UCM).
Comparison of
of wind
wind speed
speed profiles
profiles of
of remote
remote sensing
sensing
measurements
(LIDAR
Figure
Figure 11.
11. Comparison
Comparison
of
wind
speed
profiles
of
remote
sensing measurements
measurements (LIDAR
(LIDAR and
and
(WRF
and
WRF–UCM:
symbols).
SODAR:
lines)
and
numerical
weather
predictions
SODAR: lines) and numerical weather predictions (WRF and WRF–UCM: symbols).
Figure
12
presents
scatter
plots
comparing
remote
sensing
measurements
and
Figure 12
12presents
presents
scatter
plots
comparing
remote
sensing
measurements
and numerical
numerical
Figure
scatter
plots
comparing
remote
sensing
measurements
and numerical
weather
weather
prediction
during
the
campaign
period
at
400
m
aboveground.
The
wind
direction
showed
weather
prediction
during
the
campaign
period
at
400
m
aboveground.
The
wind
direction
showed
prediction during the campaign period at 400 m aboveground. The wind direction showed
22 2
goodness-of-fit
and
R
=
0.84
for
both
1-h
averaged
LIDAR
and
SODAR,
whereas
for
the
wind
goodness-of-fit and
forfor
both
1-h1-h
averaged
LIDAR
and and
SODAR,
whereas
for thefor
wind
speed
goodness-of-fit
and RR = =0.84
0.84
both
averaged
LIDAR
SODAR,
whereas
the speed
wind
2
2
this
== 0.67
for
and
SODAR,
respectively.
Through
this was
was
Rwas
0.67
and
0.61
for LIDAR
LIDAR
andand
SODAR,
respectively.
Through
this
comparison,
since
speed
thisR
R2 =and
0.670.61
and 0.61
for LIDAR
SODAR,
respectively.
Throughthis
thiscomparison,
comparison, since
since
222 = 0.67,
LIDAR
measurement
data
and
WRF–UCM
simulation
data
were
matched
at
the
level
of
R
LIDAR
measurement
data
and
WRF–UCM
simulation
data
were
matched
at
the
level
of
R
=
LIDAR measurement data and WRF–UCM simulation data were matched at the level of R = 0.67,
0.67,
WRF–UCM
data
were
considered
to
be
in
the
correction
of
remote
WRF–UCMdata
datawere
were
considered
to utilized
be utilized
utilized
inlong-term
the long-term
long-term
correction
of short-term
short-term
remote
WRF–UCM
considered
to be
in the
correction
of short-term
remote sensing
sensing
sensing measurement
measurement
data.
measurement
data. data.
Remote Sens. 2016, 8, 1019
9 of 17
Remote Sens. 2016, 8, 1019
9 of 17
Remote Sens. 2016, 8, 1019
9 of 17
(a)
(b)
Figure 12. Comparison of numerical weather prediction and remote sensing measurements for the
Figure 12. Comparison of numerical weather prediction and remote sensing measurements for the
(a) m aboveground: (a) wind speed; and (b) wind direction.
(b)
campaign period at 400
campaign period at 400 m aboveground: (a) wind speed; and (b) wind direction.
Figure 12. Comparison of numerical weather prediction and remote sensing measurements for the
campaignCorrection
period at 400 m aboveground: (a) wind speed; and (b) wind direction.
3.3. Long-Term
3.3. Long-Term Correction
Short-term data were corrected to long-term data via MCP using statistical correlation in the
Short-term
data were
corrected
long-term
data
via MCP using
correlation
in the
3.3.
Long-Termperiod
Correction
overlapped
between
the to
reference
data
(WRF–UCM)
and statistical
the short-term
LIDAR
overlapped
period between
the reference
data (WRF–UCM)
and the consistently
short-term LIDAR
measurements.
measurements.
the variance
ratio
algorithm,
which
predictions
Short-term Here,
data were
corrected
to long-term
data
viaproduced
MCP using
statisticalreliable
correlation
in the
Here,
the
variance
ratio
algorithm,
which
produced
consistently
reliable
predictions
among
the MCP
among the MCP
algorithms,
employed
[13].data (WRF–UCM) and the short-term LIDAR
overlapped
period
betweenwasthe
reference
algorithms,
was employed
[13].was extended
The wind
climate,
which
to one-year
via MCP,
at 555 m altitude
the future
measurements.
Here, the
variance
ratio algorithm,
whichdata
produced
consistently
reliableat
predictions
the
Lotte
World
Tower
is
as
follows
(see
Figure
13):
The
wind
climate,
which
was
extended
to
one-year
data
via
MCP,
at
555
m
altitude
at the future
among the MCP algorithms, was employed [13].
the•Lotte
World
Tower
iswhich
follows
(see Figure
13):
The
wind
climate,
was extended
to one-year
data via MCP, at 555 m altitude at the future
Mean
wind
speed
=as4.9
m/s;
•
•
•
•
2 (wind
the
Lotte
World
Tower
as
follows
Figure
13):
• Mean
Wind
power
density
= m/s;
160
W/m(see
class
1, poor);
wind
speed
= is4.9
•
Weibull
distribution
scale
factor
c
=
5.5
m/s,
shape
factor k = 1.72; and
2
• Wind
Mean
winddensity
speed = =4.9160
m/s;
power
W/m (wind class 1, poor);
•
Prevailing
wind
direction
WNW.
2
•
Wind power density = 160 W/m (wind class 1, poor);
Weibull distribution scale factor c = 5.5 m/s, shape factor k = 1.72; and
•
Weibull distribution scale factor c = 5.5 m/s, shape factor k = 1.72; and
Prevailing
wind direction WNW.
•
Prevailing wind direction WNW.
(a)
(b)
Figure 13. Wind climate at 555 m above ground level: (a) probability distribution of wind speed; and
(a)
(b)
(b) wind rose.
Figure 13. Wind climate at 555 m above ground level: (a) probability distribution of wind speed; and
Figure 13. Wind climate at 555 m above ground level: (a) probability distribution of wind speed;
(b) wind rose.
and (b) wind rose.
Remote Sens. 2016, 8, 1019
10 of 17
3.4. Computational Fluid Dynamics
3.4. Computational Fluid Dynamics
Figure 14 shows the comparison of simulation results of the main wind direction case, revealing
Figurewind
14 shows
thevariability
comparison
simulation
results
thevertical
main wind
direction case,
revealing
horizontal
speed
at of
the
pedestrian
levelofand
cross-sectional
wind
speed
horizontal
wind
speed
variability
at
the
pedestrian
level
and
vertical
cross-sectional
wind
variability along the main wind direction (WNW) before and after the construction of the speed
Lotte
variability
along
the
main
wind
direction
(WNW)
before
and
after
the
construction
of
the
Lotte
World Tower.
World
Tower.
The CFD analysis results considering only the Lotte World Tower and including all
The CFDbuildings
analysis results
only the
Lotte
WorldisTower
and including
all surrounding
surrounding
show considering
that the rooftop
of the
Tower
impacted
by surrounding
building
buildings
show
that
the
rooftop
of
the
Tower
is
impacted
by
surrounding
building
groups
to a low
groups to a low degree (less than 3% of wind speed), although the pedestrian level wind speeds
are
degree
(less
than
3%
of
wind
speed),
although
the
pedestrian
level
wind
speeds
are
largely
augmented
largely augmented by the Lotte World Hotel (about 75% of area; left contour plots in Figure 14). This
by the
Lotte
World
Hotel
(about
75%
of area; left
contour
in Figure
14). Thisplots
can also
be seen
can
also
be seen
in the
wind
speed
variability
at the
right plots
vertical
cross-sectional
in Figure
14.in
the wind
speed variability
thesignificantly
right verticalreduced
cross-sectional
plots inthe
Figure
14.
The
computation
load canatbe
by excluding
surrounding
buildings. As
The
computation
load
can
be
significantly
reduced
by
excluding
the
surrounding
buildings.
shown in Figure 15, only three wind directions need to be simulated using
the symmetry
of the
As
shown
in
Figure
15,
only
three
wind
directions
need
to
be
simulated
using
the
symmetry
the
cross-section (in the horizontal, vertical, and diagonal axes). With the CFD analysis on three of
wind
cross-section
(in
the
horizontal,
vertical,
and
diagonal
axes).
With
the
CFD
analysis
on
three
wind
sectors and its projection to 13 other directional sectors, individual wind maps are obtained for a
sectors
its projection
to 13
otherand
directional
sectors,using
individual
wind
maps are
obtained
total
of and
16 wind
directional
sectors
then weighted
the wind
climate
in Figure
13. for a total
of 16Figure
wind directional
sectors
and thenofweighted
thewith
wind
climate
in measurements
Figure 13.
16 shows the
evaluation
the CFDusing
results
the
LIDAR
when the
Figure
16
shows
the
evaluation
of
the
CFD
results
with
the
LIDAR
measurements
when
the good
Lotte
Lotte World Tower is not simulated. The wind speed and turbulence intensity profiles
show
World
Tower
is
not
simulated.
The
wind
speed
and
turbulence
intensity
profiles
show
good
agreement.
agreement.
(a)
(b)
Figure 14. Computation results of pedestrian-level wind speed around the Lotte World Tower:
Figure 14. Computation results of pedestrian-level wind speed around the Lotte World Tower:
horizontal
(leftcolumn);
column);and
and
vertical
(right
column)
cross-sectional
wind variability:
speed variability:
(a)
horizontal (left
vertical
(right
column)
cross-sectional
wind speed
(a) without
without
the
Tower;
and
(b)
with
the
Tower.
the Tower; and (b) with the Tower.
Remote Sens. 2016, 8, 1019
11 of 17
Remote Sens. 2016, 8, 1019
Remote Sens. 2016, 8, 1019
11 of 17
11 of 17
Figure 15. Wind directional sectors around the Lotte World Tower (top view) for wind resource
Figure 15.
15. Wind directional sectors around the Lotte World Tower (top view) for wind
Figure
mapping. Wind directional sectors around the Lotte World Tower (top view) for wind resource
resource
mapping.mapping.
(a)
(a)
(b)
(b)
Figure 16. Comparison of the CFD simulations and the LIDAR measurements at the Lotte World
Figure
16. Comparison
of the CFD
simulations
and
LIDAR measurements
at the Lotte World
Tower position:
(a) mean
speed
profiles; and
and
(b)the
turbulence
intensity profiles.
Figure
16. Comparison
of wind
the CFD
simulations
the
LIDAR
measurements
at the Lotte World Tower
Tower position: (a) mean wind speed profiles; and (b) turbulence intensity profiles.
position: (a) mean wind speed profiles; and (b) turbulence intensity profiles.
Figure 17 presents a wind resource map at 2 m above the rooftop horizontal plane of the Lotte
Figure
17 revealing
presents athe
wind
resourceofmap
2 m above
rooftop
the Lotte
World
Tower
variability
the at
annual
mean the
wind
speedhorizontal
and wind plane
powerof
density.
A
Figure
17revealing
presents athe
wind
resourceofmap
at 2 m above
the
rooftop
horizontal
plane
ofdensity.
the Lotte
World
Tower
variability
the
annual
mean
wind
speed
and
wind
power
A
low
wind
speed
area
is
identified
due
to
the
wake
from
the
square
cone
located
at
the
center
of
the
World
Tower
revealing
variability
annual
mean
wind
speedcone
and located
wind power
A low
low
wind
speed
area
isthe
identified
dueofthe
tothe
the
wake
from
the
square
atrooftop.
thedensity.
center
of
the
rooftop;
shelter
is
also
observed
from
16
columns
at
the
external
walls
in
the
Thus,
the
wind
speed
area
is
identified
due
to
the
wake
from
the
square
cone
located
at
the
center
of
the
rooftop;
rooftop;
shelter
is also
observed
from
the 16
columns
at be
theappropriate
external walls
in the rooftop.
Thus,
the
four
areas
marked
by
the
dashed
circle
are
found
to
positions
for
wind
turbine
shelter
is also
observed
fromdashed
the 16 columns
atfound
the external
walls
in the rooftop.
Thus,
the four
areas
four
areas
marked
by
the
circle
arepower
to be are
appropriate
positions
for
wind
turbine
installation
because
wind
speeds
and
wind
density
some
of
the
highest
within
a
rooftop
marked by the
dashed
circle
are found
to
be appropriate
positions
for wind
turbine
installation
installation
wind
speeds
and west–east
wind
power
density
are some
the
highest
within
a because
rooftop
area. speeds
Note because
that
the
horizontal
axes
and
south–north
areoflabeled
asarea.
x-axis
and
y-axis,
wind
and
wind
power
density
are
some
of
the
highest
within
a
rooftop
Note
that
the
area.
Note
that
the
horizontal
axes
west–east
and
south–north
are
labeled
as
x-axis
and
y-axis,
respectively.
The
z-axis
directs
vertically
upward
from
the
ground.
The
origin
is
the
center
of the
horizontal
axes
west–east
and
south–north
are
labeled
as
x-axis
and
y-axis,
respectively.
The
z-axis
respectively.
z-axis directs vertically upward from the ground. The origin is the center of the
Tower on
theThe
ground.
directs
vertically
upward
from the ground. The origin is the center of the Tower on the ground.
Tower on the ground.
Remote Sens. 2016, 8, 1019
12 of 17
Remote Sens. 2016, 8, 1019
12 of 17
Remote Sens. 2016, 8, 1019
12 of 17
(a)
(b)
(a)
(b)
Figure 17. Wind resource map at 2 m above the Lotte World Tower roof (dimensions in meters): (a)
Figure 17. Wind resource map at 2 m above the Lotte World Tower roof (dimensions in meters):
mean wind
speedresource
map; and (b) at
wind
density
map.
Figure
17.wind
Wind
mpower
above
thedensity
Lotte
World
(a) mean
speed map;map
and (b)2wind
power
map. Tower roof (dimensions in meters): (a)
mean wind speed map; and (b) wind power density map.
The BIWT examples discussed in the introduction made efforts to change the building shapes to
The BIWTspeed,
examples
discussed
inthe
the introduction
made efforts
to change
the building
shapes
to
increase
but the
cone atin
in the rooftop
of the Lotte
World
andshapes
columns
The wind
BIWT examples
discussed
thecenter
introduction
made efforts
to change
theTower
building
to
increase
wind
speed,
but
the
cone
at
the
center
in
the
rooftop
of
the
Lotte
World
Tower
and
columns
installedwind
around
the external
wallsatreduce
the wind
increase
speed,
but the cone
the center
in theinstead.
rooftop of the Lotte World Tower and columns
installed
around
the external
walls
reduce
the
wind
instead.
Figure
18
shows
the
CFD
result
for
the
main
wind
direction when changing the cross-sectional
installed around the external walls reduce the wind instead.
Figure
18
shows
the
CFD
result
for
the
main
wind
direction
changing the
cross-sectional
square
shape
the cone
of the
center
the
rooftop
a circle. when
It is noticeable
the low-wind
Figure
18of
shows
the CFD
result
forin
the
main
windtodirection
when
changing that
the cross-sectional
square
shape
of
the
cone
of
the
center
in
the
rooftop
to
a
circle.
It
is
noticeable
that
the
low-wind
speed
speed wake
decreases
changing
cross-section
to a circle.
square
shapezone
of the
cone ofdramatically
the center inwhen
the rooftop
to the
a circle.
It is noticeable
that the low-wind
wake zone decreases dramatically when changing the cross-section to a circle.
speed wake zone decreases dramatically when changing the cross-section to a circle.
(a)
(b)
(a)
(b)
Figure 18. Wind speed contours at 2 m above the Lotte World Tower roof for the main wind direction
(dimensions
in meters):
(a) square
structure;
and (b)
circular
cone
structure.
Figure
18. Wind
speed contours
at cone
2 m above
the Lotte
World
Tower
roof
for the main wind direction
Figure 18. Wind speed contours at 2 m above the Lotte World Tower roof for the main wind direction
(dimensions in meters): (a) square cone structure; and (b) circular cone structure.
(dimensions in meters): (a) square cone structure; and (b) circular cone structure.
Figure 19a presents a wind power density map when changing the cross section to circular
cone.Figure
Compared
to the square
cone
case density
in Figure
17b,when
the wake
area around
thesection
cone disappears,
19a presents
a wind
power
map
changing
the cross
to circular
19a presents
a wind
power
density
map17b,
when
changing
the around
cross section
to circular
cone.
and aFigure
relatively
high
wind
power
density
is found.
cone.
Compared
to the
square
cone
case
in
Figure
the
wake area
the cone
disappears,
Compared
to the
square
casedensity
in Figure
17b, the wake area around the cone disappears, and a
and
a relatively
high
windcone
power
is found.
relatively high wind power density is found.
Remote Sens. 2016, 8, 1019
13 of 17
Remote Sens. 2016, 8, 1019
Remote Sens. 2016, 8, 1019
13 of 17
13 of 17
(a)
(a)
(b)
(b)
Figure 19. Wind resource map at 2 m above the Lotte World Tower roof (dimensions in meters): (a)
Figure 19. Wind resource map at 2 m above the Lotte World Tower roof (dimensions in meters):
Figure
19. Wind
resource
at 2 m cone
abovestructure;
the Lotteand
World
roofmap
(dimensions
in meters):
(a)
wind power
density
map map
for circular
(b) Tower
difference
of wind power
density
(a) wind power density map for circular cone structure; and (b) difference map of wind power density
wind
power
density
map
for
circular
cone
structure;
and
(b)
difference
map
of
wind
power
density
between the square and the circular cone cases.
between the square and the circular cone cases.
between the square and the circular cone cases.
Figure 19b illustrates the ratio in wind power density between the circular and square cone
Figure
19b
inin
wind
density
between
the circular
and square
cone cases.
19b illustrates
illustrates
theratio
ratio
wind
power
density
between
the circular
and compared
square
cone
cases. In particular,
wind the
power
density
inpower
(+)
shape
layout
is significantly
improved
to
In
particular,
wind
power
density
in
(+)
shape
layout
is
significantly
improved
compared
to
(
×
)
shape
cases.
In
particular,
wind
power
density
in
(+)
shape
layout
is
significantly
improved
compared
to
(×) shape layout of wind turbines. Using this result, the (+) shape layout of wind turbines for the
layout
ofcone
wind
turbines.
Using
this Using
result, this
the (+)
shape
of wind
turbines
for the
circular
(×)
shape
layout
ofis wind
turbines.
result,
thelayout
(+) shape
layout
of wind
turbines
forcone
the
circular
case
determined.
case is
determined.
circular
cone20
case
is determined.
Figure
shows
the quantitative comparison of probability distribution of wind speed for all
Figure
shows
quantitative
comparison
of
probability
distribution
of wind
wind
speed
for all
Figure
20
shows
the quantitative
distribution
of
for
sectors between the square
and circularcomparison
cone cases. of
Forprobability
the circle case,
the frequency
of speed
wind speeds
sectors
between
the
square
and
circular
cone
cases.
For
the
circle
case,
the
frequency
of
wind
speeds
circle to the square case.
below 2 m/s decreases, whereas it increases above 7For
m/sthe
compared
m/sdecreases,
decreases,whereas
whereasititincreases
increasesabove
above77m/s
m/scompared
comparedtotothe
thesquare
squarecase.
case.
below 2 m/s
Figure 20. Probability distribution of wind speed for the square and circular cone cases.
Figure 20.
20. Probability
Probability distribution
distribution of
of wind
wind speed
speed for
for the
the square and circular cone cases.
Figure
3.5. Energy Production Calcuations
3.5. Energy
Energy Production
Production Calcuations
The reference wind turbines, 10 kW GWE-10KH (height: 4.2 m, diameter: 5.3 m) made by
The
wind
10 Korea)
kW GWE-10KH
m, diameter:
5.3
made
by
The reference
4.2
diameter:wind
5.3 m)
m)
madeand
by
Geum-Poong
Energy,
Inc.turbines,
(Jeongeup,
is selected,(height:
which is4.2
a vertical-axis
turbine,
Geum-Poong
Energy,
Inc.
(Jeongeup,
Korea)
is
selected,
which
is
a
vertical-axis
wind
turbine,
and
Geum-Poong
Energy,
(Jeongeup,
Korea)
is
selected,
which
is
vertical-axis
turbine,
Excel-S 10 kW (rotor diameter 6.7 m) made by Bergey WindPower Co. (Norman, USA), a
Excel-S
10kW
kW
(rotor
diameter
6.7
m)curves
Bergey
WindPower
Co.
(Norman,
USA),21.
a
10
(rotor
diameter
6.7
m)power
made
bymade
Bergey
WindPower
Co.
(Norman,
USA),
a horizontal-axis
horizontal-axis
wind
turbine.
The
ofby
the
two wind
turbines
are
presented
in Figure
horizontal-axis
wind
turbine.
Theofpower
curves
the twoare
wind
turbinesinare
presented
wind turbine. The
power
curves
the two
windofturbines
presented
Figure
21. in Figure 21.
Remote Sens. 2016, 8, 1019
Remote Sens. 2016, 8, 1019
14 of 17
14 of 17
Figure 21. Power curves of two small wind turbines (solid line: GWE-10KH; dashed line: Excel-S).
Figure 21. Power curves of two small wind turbines (solid line: GWE-10KH; dashed line: Excel-S).
Tables 1 and 2 summarize the annual energy production (AEP)
(AEP) and
and capacity
capacity factor
factor results,
results,
respectively, when
turbine
and
horizontal-axis
wind
turbine
are used.
NoteNote
that
respectively,
whenthe
thevertical-axis
vertical-axiswind
wind
turbine
and
horizontal-axis
wind
turbine
are used.
the calculation
assumed
100% 100%
availability
and 0%and
of all
of losses.
AEPThe
significantly
increases
that
the calculation
assumed
availability
0%type
of all
type ofThe
losses.
AEP significantly
with onlywith
a change
the cross-sectional
shape ofshape
the structure.
More specifically,
the AEPthe
forAEP
the
increases
only ainchange
in the cross-sectional
of the structure.
More specifically,
vertical-axis
wind turbine
was case
increased
to 28%, and
forand
thethat
horizontal-axis
wind turbinewind
case,
for
the vertical-axis
windcase
turbine
was increased
tothat
28%,
for the horizontal-axis
to 27%. case,
The capacity
of the vertical-axis
turbine iswind
only turbine
2/3 of that
of the
turbine
to 27%. factor
The capacity
factor of thewind
vertical-axis
is only
2/3horizontal-axis
of that of the
wind turbine. wind turbine.
horizontal-axis
The capacity factor of the simulated BIWTs can reach only 6.4%. The AEP corresponds to the
inSouth
SouthKorea).
Korea).
consumption of 17 persons (1.3 MWh/person
MWh/person in
Table 1. The AEP and capacity factor of the vertical-axis wind turbines.
Table
Wind Turbine
Wind Turbine
GWE-10KH
GWE-10KH
WT#1
WT#1
WT#2
WT#2
WT#3
WT#3
TW#4
TW#4
Overall
Overall
Square Cone Case
Square Cone Case
AEP (MWh)
Capacity Factor
AEP (MWh)
Capacity Factor
2.81
3.21%
2.81
3.21%
3.14
3.58%
3.14
3.58%
3.28
3.74%
3.28
3.74%
2.18
2.49%
2.18
2.49%
11.4
3.26%
11.4
3.26%
Circular Cone Case
Circular Cone Case
AEP (MWh)
Capacity Factor
AEP (MWh)
Capacity Factor
3.01
3.44%
3.01
3.44%
3.43
3.91%
3.43
3.91%
5.09
5.81%
5.09
5.81%
3.09
3.53%
3.09
3.53%
14.6
4.18%
14.6
4.18%
Table 2. The AEP and capacity factor of the horizontal-axis wind turbines.
Table 2. The AEP and capacity factor of the horizontal-axis wind turbines.
Square Cone Case
Circular Cone Case
Wind Turbine
Excel-S
AEP (MWh)
Capacity Factor
AEP (MWh)
Capacity Factor
Square Cone Case
Circular Cone Case
Wind Turbine
WT#1
4.53
5.17%
4.89
5.47%
Excel-S
AEP (MWh)
Capacity Factor
AEP (MWh)
Capacity Factor
WT#2
4.63
5.28%
5.12
4.84%
WT#1
4.53
5.17%
4.89
5.47%
WT#3
4.98
5.68%
7.67
8.76%
WT#2
4.63
5.28%
5.12
4.84%
TW#4
3.59
4.10%
4.77
5.45%
WT#3
4.98
5.68%
7.67
8.76%
Overall
17.7
5.06%
22.4
6.38%
TW#4
3.59
4.10%
4.77
5.45%
Overall
17.7
5.06%
22.4
6.38%
4. Discussion
The wind resource in Seoul is weak mainly due to weak synoptic conditions. According to the
4. Discussion
wind systems in the Korean Peninsula [14], a synoptic wind system is formed where WNW winds
The wind resource in Seoul is weak mainly due to weak synoptic conditions. According to the
blow to the Seoul region along Han River and northwesterly winds approximate from the west sea.
wind systems in the Korean Peninsula [14], a synoptic wind system is formed where WNW winds
As shown in the wind resource map [15] of Figure 22, as upwind in Seoul, which was wind
blow to the Seoul region along Han River and northwesterly winds approximate from the west sea.
blow from the west sea at 7 m/s or higher, was infiltrated into the inland, it was weakened into low
As shown in the wind resource map [15] of Figure 22, as upwind in Seoul, which was wind blow
wind speed due to geographical forcing caused by urban canopy in Seoul metropolitan in particular
from the west sea at 7 m/s or higher, was infiltrated into the inland, it was weakened into low wind
as well as friction with the land as shown in Figure 10. A mean wind speed and wind power density
speed due to geographical forcing caused by urban canopy in Seoul metropolitan in particular as well
Remote Sens. 2016, 8, 1019
Remote Sens. 2016, 8, 1019
15 of 17
15 of 17
as friction with the land as shown in Figure 10. A mean wind speed and wind power density were
2 at the location around the Lotte
changed
into weak
resource
as 4.1 m/s
2 at the location around the
were changed
intowind
weakpower
wind power
resource
as 4.1and
m/s108
andW/m
108 W/m
World
LotteTower.
World Tower.
Figure 22. Wind resource map of the Korean Peninsula by Korea Institute of Energy Research (mean
Figure 22. Wind resource map of the Korean Peninsula by Korea Institute of Energy Research
wind speed at 80 m aboveground).
(mean wind speed at 80 m aboveground).
5. Conclusions
5. Conclusions
Based on our findings, the BIWTs were not considered at the Lotte World Tower. However, we
Based onaour
findings,
BIWTs
wereassessment
not considered
at the Lotte
World
Tower. However,
established
procedure
forthe
wind
resource
for high-rise
BIWTs
by combining
RS, NWP,we
established
and CFD. a procedure for wind resource assessment for high-rise BIWTs by combining RS, NWP,
and CFD.
The data recovery rate of ground-based remote sensing decreased dramatically. It is very
important
to recovery
perform reconstruction
of full wind
climate
data fordecreased
at least one
year using long-term
The data
rate of ground-based
remote
sensing
dramatically.
It is very
correction
NWP and
MCP methods.
Moreover,
it is anticipated
longer
important
tovia
perform
reconstruction
of full
wind climate
data fortoatperform
least one
yearremote-sensing
using long-term
measurements
to mitigate
uncertainties
caused by seasonal
variation
wind climate.
correction
via NWP
and MCP
methods. Moreover,
it is anticipated
toofperform
longer remote-sensing
Wind
conditions
can
change
considerably
depending
on
the
building
shape,
CFD simulation is
measurements to mitigate uncertainties caused by seasonal variation of wind climate.
therefore
needed
to
analyze
the
effect
of
building
shape
accurately.
Modifying
the
building
shape is
Wind conditions can change considerably depending on the building shape, CFD
simulation
results in
up to 30%
improvement
in the
density
and
large changes
in spatialthe
variability.
therefore
needed
to analyze
the effect
of power
building
shape
accurately.
Modifying
building shape
The
wind
resource
around
the
Lotte
World
Tower
is
very
poor
despite
its
high
altitude
(555 m)
results in up to 30% improvement in the power density and large changes in spatial variability.
because it is located inland of the Korean Peninsula. However, it should be noted that the
The wind resource around the Lotte World Tower is very poor despite its high altitude (555 m)
surrounding buildings of the Lotte World Tower do not directly influence the wind speed at 555 m
because it is located inland of the Korean Peninsula. However, it should be noted that the surrounding
according to the CFD simulation.
buildings of the Lotte World Tower do not directly influence the wind speed at 555 m according to the
Changes in building shape resulted in a capacity factor of only 7%, thus the BIWTs are not
CFD simulation.
economically feasible.
Changes in building shape resulted in a capacity factor of only 7%, thus the BIWTs are not
economically feasible.
Remote Sens. 2016, 8, 1019
16 of 17
Acknowledgments: This work was conducted under the framework of Research and Development Program of
the Korea Institute of Energy Research (KIER; GP2014-0030).
Author Contributions: Hyun-Goo Kim assumed responsibility for the wind resource assessment of BIWT by
compiling the RS, NWP, and CFD results and wrote the paper; Wan-Ho Jeon contributed to CFD simulation;
and Dong-Hyeok Kim conceived and designed the LIDAR and SODAR remote-sensing campaign. All authors
read and approved the manuscript.
Conflicts of Interest: The authors declare no conflict of interest.
Abbreviations
The following abbreviations are used in this manuscript:
AEP
BAWT
BIWT
CFD
DRR
MCP
NWP
RANS
RS
WRF
UCM
Annual Energy Production
Building-Augmented Wind Turbine
Building-Integrated Wind Turbine
Computational Fluid Dynamics
Data Recovery Rate
Measure-Correlate-Predict
Numerical Weather Prediction
Reynolds-Averaged Navier-Stokes
Remote Sensing
Weather Research Forecasting
Urban Canopy Model
References
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
Rouse, R. The Pyramids of Egypt. 2016. Available online: https://public.tableau.com/en-us/s/gallery/
pyramids-egypt (accessed on 12 December 2016).
Ken, B. Sustainable Architecture and Simulation Modelling; Dublin Institute of Technology: Dublin, Ireland, 2001.
Doerr Architecture. Definition of Sustainability and the Impacts of Buildings. 2003. Available online:
http://www.doerr.org/services/sustainability.html (accessed on 12 December 2016).
Smith, R.F.; Killa, S. Bahrain World Trade Center (BWTC): The first large-scale integration of wind turbines
in a building. Struct. Des. Tall Spec. Build. 2007, 16, 429–439. [CrossRef]
Cochran, B.C.; Damiani, R.R. Integrating Wind Energy into the Design of Tall Buildings—A Case Study of
the Houston Discovery Tower. In Proceedings of the American Wind Energy Association’s WINDPOWER
2008 Conference & Exhibition, Houston, TX, USA, 1–4 June 2008.
Li, Q.S.; Shu, Z.R.; Chen, F.B. Performance assessment of tall building-integrated wind turbines for power
generation. Appl. Energy 2016, 165, 777–788. [CrossRef]
Stuart, A.; Mark, W. Strata SE1, London, UK: The First In-House Wind Generator. Struct. Eng. Int. 2011, 21,
37–40.
Chen, F.; Kusaka, H.; Bornstein, R.L.; Ching, J.; Grimmond, C.S.B.; Grossman-Clarke, S.; Loridan, T.;
Manning, K.W.; Martilli, A.; Miao, S.; et al. The integrated WRF/urban modelling system: Development,
evaluation, and applications to urban environmental problems. Int. J. Climatol. 2011, 31, 273–288. [CrossRef]
Macdonald, R.W.; Griffiths, R.F.; Hall, D.J. An improved method for estimation of surface roughness of
obstacle arrays. Atmos. Environ. 1998, 32, 1857–1864. [CrossRef]
Kato, M.; Launder, B.E. The modeling of turbulence flow around stationary and vibrating square cylinders.
In Proceedings of the 9th Symposium on Turbulence Shear Flows, Kyoto, Japan, 16–18 August 1993.
Lien, F.S.; Yee, E.; Cheng, Y. Simulation of mean flow and turbulence over a 2D building array using
high-resolution CFD and a distributed drag force approach. J. Wind Eng. Ind. Aerodyn. 2004, 92, 117–158.
[CrossRef]
Gryning, S.E.; Floors, R.R.; Peña, A.; Batchvarova, E.; Brümmer, B. Weibull wind-speed distribution
parameters derived from a combination of wind-lidar and tall-mast measurements over land, coastal
and marine sites. Bound. Layer Meteorol. 2016, 159, 329–348. [CrossRef]
Rogers, A.L.; Rogers, J.W.; Manwell, J.F. Comparison of the performance of four measure-correlate-predict
algorithms. J. Wind Eng. Ind. Aerodyn. 2005, 93, 243–264. [CrossRef]
Remote Sens. 2016, 8, 1019
14.
15.
17 of 17
Kim, J.S.; Kim, H.G.; Park, H.D. Surface wind regionalization based on similarity of time-series wind vectors.
Asian J. Atmos. Environ. 2016, 10, 80–89. [CrossRef]
Korea Institute of Energy Research. Korea New and Renewable Energy Resource Atlas, 2nd ed.; Ministry of
Science, ICT and Future Planning, New and Renewable Energy Resource Center: Daejeon, Korea, 2015.
Available online: http://www.kier-solar.org/pdf/getFile.do?type=WIND (accessed on 12 December 2016).
© 2016 by the authors; licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons Attribution
(CC-BY) license (http://creativecommons.org/licenses/by/4.0/).