System and Species Selection

Vertical Greening in a Subtropical City

LM Chu, PhD School of Life Sciences, CUHK Email: leemanchu@cuhk.edu.hk

First International Conference on Green Walls Meeting the Challenge of a Sustainable Urban Future: The Contribution of Green Walls Staffordshire University, Stoke-on-Trent, UK 4-5 September 2014

Substrate-based

Major types of VGS according to planter position and orientation

Green façade Living wall

Near-wall planting Hanging Hydroponics

Hydrophilic bags

Wire

Vertical Angled Horizontal

1.1 Classification Vertical Greenery System (VGS)

Direct Indirect

Climbing

2

1. VERTICAL GREENING

Substrate-based

Direct Indirect

Major types of VGS according to planter position and orientation

Support type Carrier type

Near-wall planting Hanging Hydroponics

Hydrophilic bags

Wire

Vertical Angled Horizontal

Vertical Greenery System (VGS)

Climbing

3

1.1 Classification

Number of SCI publications (N=49) on vertical greening from 2000 to 2014 (June) classified by greenery systems and research topics.

1.3 Research on vertical greening

4

62% 13%

7%

4% 5%

5%

2% 2% Thermal behavior / MicroclimatesBuilding energyAir purificationCost-benefit analysis / Life-cycle analysisAcousticsSystem water usagePublic perceptionPlant growth performance

02468

10121416

Support-type systems

Carrier-type systems

Both systems

Unknown systems

2. FUNCTIONS AND BENEFITS 2.1 Why do we need vertical greening? Lack of greening space

5

Area of open spaces + parks

Mumbai 1.1 m2/person

Hong Kong 3.5 m2/person

Singapore 4.0 m2/person

New York 26.4 m2/person

London 31.7 m2/person

Increase green plot ratio

Horizontal planting vs Vertical planting

6

Indoor 36% Outdoor

64%

Proportion of outdoor & indoor VG projects in Hong Kong.

0102030405060708090

Number of VG projects in Hong Kong.

7

2. FUNCTIONS AND BENEFITS 2.1 Why do we need vertical greening? • Aesthetic appeal

• Urban greening (linking buildings and landscaping) • alleviate visual barrier in urban canyon • return nature to urban habitats

8

0

20

40

60

80

鵝掌柴

吊蘭

虎尾蘭

蚌花

黄金葛

腎蕨

天冬

紅掌/粉掌

灑金榕

椒草

Fre

quen

cy

Outdoor VGS

Sche

ffler

a sp

p.

Chlo

roph

ytum

co

mos

um

Sans

evie

ria sp

p.

Trad

esca

ntia

spat

hace

a 'C

ompa

cta'

Nep

hrol

epis

spp.

Epip

rem

num

pin

natu

m

Aspa

ragu

s den

siflo

rus

Anth

uriu

m

andr

aean

um

Codi

aeum

spp.

Pepe

rom

ia sp

p.

0

20

40

60

80

鵝掌柴

富貴竹/

竹蕉

黄金葛

袖珍椰子

灑金榕

吊蘭

椒草

虎尾蘭

印度橡膠榕

腎蕨

Fre

quen

cy Indoor VGS

Sche

ffler

a sp

p.

Chlo

roph

ytum

co

mos

um

Epi

prem

num

pin

natu

m

Codi

aeum

spp.

Sans

evie

ria sp

p.

Pepe

rom

ia sp

p.

Nep

hrol

epis

spp.

Drac

aena

spp.

Cham

aedo

rea

eleg

ans

Ficu

s ela

stic

a 'B

urgu

ndy’

Epi

prem

num

pin

natu

m

Epi

prem

num

pin

natu

m

Plant species most commonly used for outdoor & indoor VGS in Hong Kong. 9

Dracaena spp.

Epipremnum pinnatum

Chamaedorea elegans

Codiaeum variegatum

Indoor

Schefflera spp.

Outdoor

Schefflera spp.

Chlorophytum

comosum

Sansevieria spp.

Tradescantia spathacea 'Compacta'

Epipremnum pinnatum

Most commonly used plant species for outdoor & indoor VGS.

10

2. FUNCTIONS AND BENEFITS 2.1 Why do we need vertical greening? • Ecosystem services

• heat mitigation • dust removal • stormwater retention • noise reduction • biodiversity support

11 Buttercups, crocuses, strawberries & geraniums on the VG in Victoria, London.

VG in Edgware Road Tube Station, London.

cc31

Typewritten Text

International Green Wall Conference - 4-5th September 2014 - The Green Wall Centre, Staffordshire University, Stoke-on-Trent, UK

3. RESEARCH ON VERTICAL GREENING 3.1 Heat reduction ability of planter-on-ground type VGS

13

Experimental setup Outdoor climber system on east-facing wall 13 species on planters and metal mesh Design and installation of modular planter

Average maximum temperature reduction ability of 13 vine species under different weather conditions.

14

0

1

2

3

4

Pyro

steg

ia v

enus

ta

Boug

ainv

illea

spec

tabi

lis

Qui

squa

lis in

dica

Pand

orea

jasm

inoi

des

Ure

chite

s lut

ea

Cler

oden

drum

thom

sona

e

Clito

ria te

rnat

ea

Loni

cera

japo

nica

Pass

iflor

a ed

ulis

Man

devi

lla ×

am

abili

s

Thun

berg

ia g

rand

iflor

a

Man

soa

allia

cea

Bauh

inia

gla

uca

Tem

pera

ture

redu

ctio

n (o C

) Clear sunny

Sunny intervals

Cloudy

Mean hourly temperature behind canopy for Bauhinia glauca (top) and Pyrostegia venusta (bottom).

15

20

25

30

35

40

45

22 24 26 28 30 01 03 05 07 09 11 13 15 17 19 21 23 25 27 29 31 02 04 06 08 10 12 14 16 18 20 22 24 26 28 30

Tem

pera

ture

(o C)

Control Bauhinia glauca

20

25

30

35

40

45

22 24 26 28 30 01 03 05 07 09 11 13 15 17 19 21 23 25 27 29 31 02 04 06 08 10 12 14 16 18 20 22 24 26 28 30

Tem

pera

ture

(o C)

Control Pyrostegia venusta

Jul Aug Sep

Jul Aug Sep

Diurnal temperature changes behind canopy.

Bauhinia glauca

Pyrostegia venusta

24

26

28

30

32

34

36

38

40

00:00 04:00 08:00 12:00 16:00 20:00 00:00

Tem

pera

ture

(o C)

24

26

28

30

32

34

36

38

40

00:00 04:00 08:00 12:00 16:00 20:00 00:00

Tem

pera

ture

(o C)

16

Pearson correlation between heat reduction behind canopy and canopy characteristics on clear sunny days.

Correlation coefficient

Leaf number 0.26 Leaf density (no./cm3) 0.04 Leaf angle (o) 0.04 Effective LAI 0.42 Leaf area (cm2) -0.10 Canopy thickness (cm) 0.62* Canopy coverage (%) 0.84** LAI 0.73** Leaf thickness (cm) 0.21 Leaf color (hue) 0.11 Leaf length to width ratio 0.10 PC 1 0.16 PC 2 0.79** PC 3 -0.07 PC 4 -0.23 PC 5 0.38 * p < 0.05; ** p < 0.01 17

Canopy characteristics PC1 PC2 PC3 PC4 PC5

Leaf number -0.79 0.54 0.27 0.02 -0.10

Leaf density (no./cm3) -0.82 0.34 0.37 0.18 -0.13

Leaf angle (o) -0.73 0.04 -0.44 0.01 0.43

Effective LAI 0.80 0.45 0.31 0.07 -0.19

Leaf area (cm2) 0.77 -0.41 -0.16 -0.33 0.11

Canopy thickness (cm) -0.06 0.82 -0.46 -0.07 0.11

Canopy coverage (%) 0.18 0.84 -0.35 -0.04 -0.14

LAI 0.56 0.73 0.10 -0.04 0.26

Leaf thickness (cm) 0.31 -0.14 0.55 -0.27 0.35

Leaf color (hue) -0.44 0.17 0.34 -0.67 -0.13

Leaf length:Leaf width 0.29 0.16 -0.18 -0.56 -0.53

Variance explained (%) 35 29 11 8 6

Major PCs and their attributes of canopy characteristics of the 13 climbing species.

18

• There were difference in thermal reduction performance between the various species studied.

• Pyrostegia venusta was significantly higher than Bauhinia glauca in their temperature reduction ability in the support-type vertical greening system.

• Thermal performance was significantly related to canopy attributes such as canopy thickness, canopy coverage & LAI.

• Denser, thicker and better-covered vegetation was better in terms of microclimate modification.

Summary

3.2 Heat reduction ability of modular VGS modular carrier system outdoor east-facing wall 2 species

Ophiopogon japonicus cv Nanus Nephrolepis exaltata

on modular panels

19 Experimental setup

On substrate surface (behind canopy)

Wall

Substrate panel

On wall surface (behind panel)

12 cm

Temperature sensor

42 cm

Modular panel

Ophiopogon japonicus Nephrolepis exaltata

Plant height (cm) 6.02 16.8 Canopy thickness (cm) 4.91 14.5 Canopy coverage (%) 55.0 71.2 LAI 0.63 1.10

Canopy characteristics of Ophiopogon japonicus cv Nanus & Nephrolepis exaltata on modular VGS.

20

Mean hourly temperature on the substrate surface for Ophiopogon japonicus cv Nanus (top) & Nephrolepis exaltata (bottom).

21

20

25

30

35

40

28 29 30 31 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26

Tem

pera

ture

(o C) Control Liriope spicata

20

25

30

35

40

28 29 30 31 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26

Tem

pera

ture

(o C)

Control Nephrolepis exaltata

Ophiopogon japonicus

Aug Sep

Aug Sep

22

24

26

28

30

32

34

36

00:00 04:00 08:00 12:00 16:00 20:00 00:00

Tem

pera

ture

(o C)

22

24

26

28

30

32

34

36

00:00 04:00 08:00 12:00 16:00 20:00 00:00

22

24

26

28

30

32

34

36

00:00 04:00 08:00 12:00 16:00 20:00 00:00

22

24

26

28

30

32

34

36

00:00 04:00 08:00 12:00 16:00 20:00 00:00

Tem

pera

ture

(o C)

Diurnal temperature changes behind canopy (on substrate surface) & panel (on wall surface).

Nep

hrol

epis

exal

tata

O

phio

pogo

n ja

poni

cus

Substrate surface Wall surface

22

Mean temperature reduction of Nephrolepis exaltata (blue) & Ophiopogon japonicus cv Nanus (red) under different weather conditions (* p<0.05).

23

24

• Nephrolepis exaltata had significantly higher temperature reduction ability than Ophiopogon japonicus cv Nanus in carrier-type VGS.

• This was attributed to its denser coverage & taller individuals.

• Nephrolepis exaltata reduced 1.5oC behind canopy while Ophiopogon japonicus cv Nanus was ineffective.

• Temperature reduction on the wall surface was negligible with or without vegetation.

• Species selection for carrier type vertical greening in maximizing temperature reduction efficiency would not be significant when there was thick substrate (thermal mass) with high in situ water content (thermal buffer).

Summary

3.3 Performance comparison between support-type and carrier-type VGS

• Species with best temperature reduction ability - Pyrostegia venusta & Nephrolepis exaltata

• Focused on wall surface temperature reduction which affects cooling energy loads & indoor living comfort

25

Temperature sensors

Air- canopy interface

Behind canopy

Wall surface

3 cm

10 cm 3 cm

Planting mesh

Air-canopy interface

Substrate surface

Wall surface

Substrate panel

12 cm 10 cm

20

22

24

26

28

30

32

34

36

00:00 06:00 12:00 18:00 00:00 06:00 12:00 18:00 00:00 06:00 12:00 18:00 00:00

Tem

pera

ture

(o C)

Air-canopy interfaceBehind canopyWall surface

20

22

24

26

28

30

32

34

36

00:00 06:00 12:00 18:00 00:00 06:00 12:00 18:00 00:00 06:00 12:00 18:00 00:00

Tem

pera

ture

(o C)

Air-canopy interfaceSubstrate surfaceWall surface

Diurnal temperature changes at different locations. 26

Nephrolepis exaltata

Pyrostegia venusta

Average daily maximum temperature of bare wall surface, ambient air & different locations of the support-type system & carrier-type system.

aOne-way ANOVA followed by Tukey HSD post-hoc test to determine any significant difference between mean daily maximum wall surface temperature

Daily maximum temperature (oC) a

Wall surface temperature reduction (oC)

Bare wall surface 35.2 a

Ambient air 31.9

Support system

Air-canopy interface 34.9

Air cavity 33.9

Wall surface 31.3 b 3.8

Modular system

Air-canopy interface 34.0

Substrate surface 32.4

Wall surface 28.8 c 6.4

27

Wall surface temperature under two types of VGS & a bare wall.

28

22

24

26

28

30

32

34

36

06:00 12:00 18:00 00:00 06:00

Tem

pera

ture

(o C)

Bare

Support-type

Carrier-type

29

• Support-type VGS was less effective than carrier system in heat reduction at the wall surface.

• There was a lag in heat transfer in the wall of carrier system, which resulted in slower cooling in late afternoon & higher temperature than bare wall after sunset.

• A higher continuous inflow of heat at nighttime would likely to cause excessive air-conditioning or discomfort to residents in hot summer nights.

• Although carrier-type vertical greenery system did better in heat reduction, it should be carefully selected for its installation in residential buildings.

Summary

Floor plan of experimental flats

Window opened for natural ventilation

Window sealed

Control (no VGS)

With VGS

Sensors

GW-VG1

2 flats on 4/F & 5/F at a construction site SW facing wall GW-VG1 with Peperomia claviformis summer 2013 – spring 2014 thermal performance of naturally ventilated flats with

& without GW-VG1

30 3D view of experimental flats

3.3 VGS on human thermal comfort

Sensors and measurements • temperature sensor (DS18B20 1-wire digital thermometer,

Dallas Semiconductor, USA) • air humidity sensor (HIH-4000-003, Honeywell Sensing and

Control, USA) • air velocity transducer (8455, Trust Science Innovation (TSI),

USA) • signal convertor (Ethernet 1-wire host adapters, Embedded

Data Systems, USA) • automatically logged in a computer every 30 min

radiant, air & surface temperatures; solar & infrared radiation; heat flux & wind speeds were measured

Predicted mean vote (PMV) index PMV based on heat balance principle & experimental data (thermal variables) collected Personal factors

metabolic rate of the subject clothing insulation

Environmental factors air temperature surrounding surfaces temperature (radiant

temperature) water vapor pressure relative air velocity

Predicted percentage of dissatisfied (PPD) index • PPD predicts the percentage of a large group

of people likely to feel ‘too warm’ or ‘too cool’ • ISO 7730 • Indoor

Description Numerical value* Very hot > +3 Hot +3 Warm +2 Slightly warm +1 Neutral 0 Slightly cool -1 Cool -2 Cold -3 Very cold < -3

PMV comfort scale (ASHRAE)

* Numerical value usually ranged between - 3 to + 3. Thermal neutrality when heat generated by human metabolism is allowed to dissipate to maintain thermal equilibrium/comfort.

2 indices to show indoor thermal comfort

-8

-6

-4

-2

0

2

4

6

8PM

V

0

20

40

60

80

100

PPD

With VG Control

Rainy Hot and hazy

Hot and sunny

Summer

Standing at rest (1.2 met) Clothing insulation (0.28) • Panties ; sleeveless shirt; short sleeve shirts;

short trousers; socks; thin soled shoes 33

Control With VGS

PMV exceeded +5 at around 12:00 to 16:00 under hot & sunny days in flats without VGS

PPD reached 100 for control; PPD was 60-70 with VGS

30% more people felt discomfort after 18:00

34

• VGS was effective in producing better indoor thermal comfort for residence under summer condition as suggested by the near neutral PMV.

• VGS gave lower PPD during daytime than bare wall. • There were more dissatisfaction (30% more people felt

discomfort) in flat with VGS after 18:00 under natural ventilation.

• VGS more recommended for commercial buildings where most activities were carried out in daytime.

Summary

Floor plan of experimental flats

Window sealed

Control (no VGS)

With VGS

PMV sensors

GW-VG1

3.4 Urban C footprint of a modular VGS same 2 flats as for 3.3 LCA (GaBi Software)

35 3D view of experimental flats

y = 0.1572x - 3.3072 R² = 0.5908 P < 0.001

0.0

0.4

0.8

1.2

1.6

2.0

23 24 25 26 27 28 29 30 31

Elec

trici

ty s

avin

g (k

Wh/

day)

Temperature (oC)

Regression between daily average temperature & energy saving of GW-VG1 in August and September 2013.

36

Cooling potential of VGS (8.22 m2) Estimated energy saving = 63 kWh/y

0%

20%

40%

60%

80%

100%

Glo

bal w

arm

ing

pote

ntia

l10

0 ye

ars

Aci

dific

atio

n po

tent

ial

Eutro

phic

atio

n po

tent

ial

Ozo

ne la

yer d

eple

tion

pote

ntia

l

Abi

otic

dep

letio

nel

emen

ts

Abi

otic

dep

letio

n fo

ssil

Fres

hwat

er a

quat

icec

otox

icity

pot

entia

l

Hum

an to

xici

ty p

oten

tial

Mar

ine

aqua

ticec

otox

icity

pot

entia

l

Phot

oche

mis

try. o

zone

crea

tion

pote

ntia

l

Terr

estri

c ec

otox

icity

pote

ntia

l

Materials Transportation Usage

Relative contribution of stages on the environmental burdens of VGS over a 50 year life span.

37

Impact category Impact indicator VG environmental burdens

Environmental impacts avoided by electricity saving

Payback period (y)

Marine aquatic ecotoxicity potential

kg DCB equivalent

15600 979 15.9

Abiotic depletion fossil MJ 1510 78.6 19.3 Global warming potential 100 years

kg CO2 equivalent

144 6.73 21.5

Human toxicity potential kg DCB equivalent

4.88 1.98 2.47

Acidification potential kg SO2 equivalent

0.25 0.04 6.84

Terrestrial ecotoxicity potential

kg DCB equivalent

0.24 0.01 24.0

Freshwater aquatic ecotoxicity potential

kg DCB equivalents

0.23 0.06 3.83

Eutrophication potential kg phosphate equivalent

0.04 4.46E-3 7.85

Photochemistry - ozone creation potential

kg ethene equivalent

0.03 3.16E-3 8.86

Payback period of VGS (per m2 basis) estimated using 11 environmental impact indicators in LCA.

38

Payback time (offset C debt) = 22 y

39

• VGS on wall could save as much as 16% of electricity consumption for air-conditioning in August and September (typical hot and wet summer months), which was about 63.8 kWh.

• According to LCA, the material manufacturing stage contributed to over 60% of all the environmental impact categories.

• Based on comparison between the environmental burdens & benefits for cooling (50-y building service life), environmental burdens of VGS on marine aquatic ecotoxicity potential, abiotic depletion of fossil fuels and global warming potential could be paid back in 22 years.

Summary

40

• Plant species choice is important for VGS with respect to growth & thermal performance, esp. for support system.

• Canopy features were major attributes to heat reduction. • Carrier VGS was capable of reducing heat flux through building

wall. Hence decreasing interior cooling energy consumption. • Carrier VGS gave generally better feeling of indoor satisfaction

especially under hot & sunny days as indicated by PMV & PPD indices. This resulted in better thermal comfort for occupants staying in rooms under summer condition.

• There was a lag in heat flux through the wall, which resulted in prolonged heat after sunset. This may increase cooling energy use for residential premises.

• The payback time could be long, but the benefits of the VGSs in C sequestration & energy saving should not be overlooked.

4. CONCLUSIONS

Checklist for successful VGS Purpose Aesthetics Ecosystem services Site Location (Outdoor/Indoor) Orientation Building height Climate Loading Scale System Type Irrigation Substrate Plant Growth habit/requirement Function/Purpose Pure/Mixed culture Maintenance Pest/Weed control Fertilizer Hardware repair Plant replacement

41

Bosco Verticale, Milan (2013) Clearpoint Residencies, Sri Lanka (2015)

Vegitecture Urban Greening for the Future?

Architecture for the Future?

42

VGS for Sustainable Built Environment?

ACKNOWLEDGEMENTS • Environment and Conservation Fund

• Everplant Technology Ltd

• GreenWalls Bioengineering (HK) Ltd

• Hong Kong Greenlink Kusters Co Ltd

• Strongly International Ltd

• Poying LAI (MPhil student & research assistant)

• Sharon CHAN (Research assistant)

• Lan PAN (PhD student)

43