ventilation processes from building to city scale

ASI 1 ‐ Prof. Janet Barlow

1

The Breathing City – ventilation processes from building to city scale

Prof Janet F. BarlowProfessor in Environmental Physics Dept. of Meteorology, Uni. of Reading, UK

Thanks to:Ewan O’Connor (lidar expertise)Rosy Wilson, John Lally (technical support)Westminster Council (Steve Neville), BT


2

3

Fūjin, Shinto God of The Wind

4

fēng shuǐ

“wind water”


3

Human ventilation

Building ventilation

Urban microclimates impacting building ventilation

Street to neighbourhood scale pollution dispersion

Urban boundary layer ventilation

Thermally driven flows – river, slope, sea breeze

Overview

London‐based urban research projects

DAPPLE (2002‐2009) – street level dispersion

(REPARTEE (2006‐2007) – vertical pollutant distribution)

ACTUAL (2009‐2014) – building design interactions with urban climate at a range of scales

ClearfLo (2010‐2013) – air quality at city scale

Refresh (2013‐2018) – indoor thermal comfort and human performance


4

Thanks to Steve Evans, UCL

Regents Park

BT TowerRoof-top site

1.6 km

BT Tower at190 m

Roof top at ~20m

urban sodar<200m

Doppler lidar90m < z < 2000m

ACTUAL: Observing flow at range of scales

Barlow et al. 2009 BLM

Wood et al. 2010 BLM

Helfter et al. 2011 ACP

Barlow et al. 2011 JWEIA

Bradley et al. (2016) Met. Zeitschrift

Wood et al. 2012 STOTEN

Pauscher et al. ICUC8 poster (KCL)

KCL: Scintillometers(various heights)

Lane et al. 2013 JWEIA

Drew et al. 2013 JWEIA


5

Human ventilation

The breathing zone

• Movie – Schlieren photography of human plume

Thanks to Dr Julian Tang and Prof Gary Settles, via Prof Cath Noakes


6


12

Heat flows in a building

Hacker et al., (2005) Beating the Heat: Keeping UK Buildings Cool in a Warming Climate, UKCIP Briefing report


7

Types of building ventilation

resaveguide.lbl.gov, Lawrence Berkeley National Lab, US

• Mechanically ventilated

See also Awbi, H. (2003) Ventilation of Buildings


Linden (1999) Ann. Rev. Flu. Mech.

• Naturally ventilated

Mixing ventilation Displacement ventilation


8


Center for the Built Environment, UC Berkeley, www.cbe.berkeley.edu

• Mixed mode

Infiltration

www.clemdesign.com

• Air leakage through cracks and gaps

• IR camera showing cold air infiltration beneath window


9

Natural ventilation of urban buildings

• How is ventilation of a building changed by presence of other buildings?- flow pattern, turbulence,temperature, inside and out

?

Wind speed UH

eigh

t z

School of Slavonic and East European Studies, UCL, London


10

Human breathing penetrates the “thermal plume” that we carry around with us

Building ventilation can be natural, mechanical or mixed mode ventilation

Human plumes can influence room ventilation!

Urban microclimate can change effectiveness of naturally ventilated or mixed mode systems

Insights from research

Human and building ventilation

Human ventilation






Overview


11

How do urban winds and temperatures drive infiltration of buildings?

Aidan Brocklehurst, PhDwith Building Research Establishment (BRE)The ACTUAL project

• Gas measurements by Building Research Establishment (BRE)• Experimental period: 24.11.11 to 13.1.12; winter, less buoyant flow• Opportunity! Empty, unheated building – examine physical response


12

Indoor and outdoor pollution concentrations

• Lag of indoor concentrations ~ 2 hours• Relate to infiltration rate• Determine dependence of infiltration rate

on wind direction, speed

Northerly winds, high concentration in courtyard

Daily Infiltration Rates• MIAQ: Multizone Indoor Air

Quality method• EnergyPlus infiltration

model = f(U)


13

Can urban canopy flow impacts on building natural ventilation rates be quantified?

The Refresh Project (2013-2018)

Natural ventilation of urban buildings (the Refresh project)

• Up to 90% of our time spent indoors (Leech et al. 2002)• UK building stock relies heavily on natural ventilation (Meijer et

al, 2009)• Work hours will be lost due to overheating of cities under

climate change (Greater London Authority, 2006)• How well can we predict natural ventilation of buildings in a

city?


14

Ventilation measurements using the Silsoecube

• Site and building behaviours are well known in wind engineering (Richards 2012; Straw 2000; Yang 2004) – rural site in UK

• Previous experiments only undertaken during certain conditions• Complete control of instrument positions and ventilation set-up • Aiming to gain an understanding of how an urban

canopy may affect natural ventilation rates in building

286 m6 m

Measuring:RainfallRadiationWind speed Wind directionTemperaturePressureCO2 concentration

6 m

Ref wind at 10, 6 m


15

Measuring ventilation rates• Wind, met and pressure data September 2014 to July 2015• 156 CO2 tracer gas decay releases undertaken (Sherman,

1990):

FILL DECAY

ln (Ci(t)-Ca) = ln(C0- Ca) – λTtλT = 5.76 ACH 0.01

C0

Ca


16

Preliminary results• A limited array of buildings reduces

cross ventilation rate by c. 28% single sided ventilation rate by c. 64%

• Relatively small shifts in wind direction can reduce ventilation rate by 30 to 35%

• Single sided ventilation in a building at the edge of an array appears to be the same as an isolated building

www.refresh-project.org.uk

Human ventilation






Overview


17

Can the design and layout of buildings improve street-level environmental quality?

Salfex 2002; windtunnel experiments, 2002-2013

Street canyon flow Define street aspect ratioHeight /widthH/W

Isolated roughnessH/W < 0.3

Wake interference0.3 < H/W < 0.6

Skimming flowH/W > 0.6


18

Salfex 2002

Salford, Greater Manchester

Street level flowflat roof H/W=1.0

Pitched roof H/W=1.0

• Street width controls flux of heat and pollution to air above

• Adding pitched roof can reduce flux by up to 40%

Barlow and Belcher (2002)

Harman, Barlow and Belcher (2004)

Pascheke, Barlow and Robins (2008)


19

What is the effect of varying roof height on pollutant concentrations?• Cheng and Castro (2002)

F=0.25 cuboid arrays• Height is “uniform, U” or “non-

uniform, NU” – normal distribution of roof heights about mean

U NUu/Uref 0.058 0.063 +9%UH/Uref 0.22 0.18 -18%

NU: vertical mixing betterhorizontal advection worse

Pascheke et al. 2008 Boundary-Layer Meteorology


20

Pascheke et al. 2008 Boundary-Layer Meteorology

What are the ventilation pathways around intersections?

DAPPLE projects, 2002-2009


21

2004 campaign BT Tower NE wind direction

Roof-top references UH Led by A. Robins

• DAPPLE-HO –emergency response tracer experiments 2007/8

• Near field dispersion

• Dosage:

• “channelling” – a bluff element canopy phenomenon?Wood et al. 2009, BAMS

∗

2


22

Street canyon flow is relatively straight-forward

Intersection flow is complicated!!

Wood C.R. et al. (2009) Bulletin American Meteorol. Soc.


23

QUESTION: is this a street canyon? What are the appropriate archetypes/typologies?

• Tall building urban canopy hard to study!

• Numerical model: Large Eddy Simulation

• (PALM, Siegfried Raasch, Marcus Letztel, Leibniz Uni. of Hannover)

• Movie 1: Kowloon• Movie 2: Macau

Human ventilation






Overview


24

The Boundary Layer• Bottom layer of the atmosphere, usually around 1 km deep, but in mid‐latitudes can vary from 100 m to 3 km.

• The surface influences the ABL by friction and by heat fluxes at the ground.

• Temperatures and other quantities vary diurnally, unlike the free atmosphere above.

• Characterised by turbulence, which is generated by wind shear (wind‐speed is zero at the surface). Temperature gradients can either generate or suppress turbulence.

• Boundary layer clouds: predominantly fair‐weather cumulus, stratocumulus and fog.

Oke T.R., Boundary Layer Climates; Stull R.B. The Atmospheric Boundary Layer

Boundary layer structure

48


25

Boundary layer structure

49

Boundary layer evolution

• Surface energy balance (QH) determines evolution

QH positive – convectiveQH negative – stableQH ~ 0 - neutral

50


26

~2-5H

~0.1zi

z

zi~1km

windspeed potentialtemperature

The Urban Boundary Layer

Building, H

Street canyon, H/W

Surface layer, 0.1zI

Urban Boundary Layer, zi

Street 10-100m Neighbourhood 100-1000m City 10-20km

Person 1m

Barlow, J.F. (2014) Progress in observing and modelling the urban boundary layer, Urban Climate, 10, 216-240

From Fernando (2010), Annual Review of Fluid Mechanics


27

Convective UBL with zero wind‐speed

• Higher heat flux over city causes mesoscalethermal circulation

• Similar to a lake/sea breeze:- Hot air rises over city, causing a low pressure

anomaly at surface- Horizontal pressure gradient “sucks” in cooler, rural

air at low levels

53

Convective UBL with non‐zero wind

From Oke 198754

• Synoptically driven winds encounter change in roughness, heat flux

Gradual deceleration of wind; adjustment of surface fluxes; adjustment of mean profiles of wind and temperature

internal boundary layer is formed


28

Atmospheric boundary layers exist in response to the Earth’s surface – exchange of momentum, heat, moisture, gases, particulates, etc.

Urban boundary layers have distinctive characteristics due to the large roughness elements and heterogeneous surface types

Large contrasts exist between rural and urban boundary layers – exchange of air = ventilation

insights from research

Boundary layers

How differently do the urban and nearby rural boundary layers evolve?

Sian Lane, PhD Christos Halios, Dan Drew, post-docsREPARTEE project 2006-2007ACTUAL project, 2009-2014Clearflo project, 2010-2014


29

Ground-based remote sensing

The ACTUAL project www.actual.ac.uk

• Doppler lidar (Halo Photonics) in London and Chilbolton

• Eye-safe (1.5 μm)

• Gate length 30m (36m Ch)

• Integration time: 3.6s (40s Ch)

• Scan pattern (L): vertical stare, Doppler Beam Swinging

Harwell

Chilbolton

Detling

• 4 components of net radiation Kipp and Zonen CNR1 (shortwave radiation)


30

Lidar observations – 6th Nov 2007, convective conditions

Barlow et al. 2011, Atmospheric Chemistry and Physics, 11 (5), 2111-2125

High pressure system causes stable profiles

• Cloudless case study period 28.09.11 to 2.10.11


31

28.09 – 02.10.11 case study: mixing height trends for urban and rural sites

• Derived using simple threshold• MH = height up to which vertical velocity variance w

2 > 0.1 m2s-2

Barlow et al. 2014, Environmental Fluid Mechanics, 15 (2), 373-398

Comparison of mixing heights (monthly means 2011)

2 4 6 8 10 12 14 16 18 20 22 24

Hour (UTC)

2

4

6

8

10

12

Mo

nth

0

200

400

600

800

1000

1200

1400

1600

mix

ing

he

ight

(m

)

2 4 6 8 10 12 14 16 18 20 22 24

Hour (UTC)

2

4

6

8

10

12

Mo

nth

0

200

400

600

800

1000

1200

1400

1600

1800

mix

ing

hei

gh

t (m

)

Rural site (Chilbolton)

Urban site (London)


32

Comparison of mixing heights (monthly means 2011)

2 4 6 8 10 12 14 16 18 20 22 24

Hour (UTC)

2

4

6

8

10

12

Mo

nth

0

200

400

600

800

1000

1200

1400

1600

mix

ing

hei

gh

t (m

)

2 4 6 8 10 12 14 16 18 20 22 24

Hour (UTC)

2

4

6

8

10

12

Mo

nth

-200

0

200

400

600m

Rural site (Chilbolton)

MH difference(urban-rural)

Human ventilation






Overview


33

Dan Drew, post-doc

• Doppler lidar observed wind-speed using Doppler beam swinging (DBS)

• Comparison with BT winds: Ulidar=0.98 UBT + 0.56

• Sampling error ~ 0.4 ms-1

L

LiDAR gate resolution

Lane S.E. et al., (2013), JWEIA

How well do Doppler lidar measured wind-speeds compare with mast-based measurements?

What is the evolution of winds along an urban river throughout a day?

Wood et al. 2012, STOTENCurtis Wood, post-doc (now FMI)

Collaboration with Sue Grimmond, now at Reading

• C – cloudy, S – sunny• Feb to May 2011

• New horizontal scanning lidar technique

• Comparison with sonic anemometry, scintillometer

411


34

Diurnal cycle in wind speed – vertical mixing

• Time series of wind-speed at urban site (UWCC), rural site (Uch) and for selected lidar gate heights.


Diurnal cycle in wind direction – downslope flows

• Time series of wind direction at urban site (phi_WCC) and for selected lidargate heights.


DAY

NIGHT


35

+

• UK Met Office UKV model(operational weather forecast)• 0500UTC wind vectors• Weak, NE, downslope flow,

c. 100 m deep


36

Urban areas in complex terrain

From Fernando (2010), Annual Review of Fluid Mechanics

Simplifying complex terrain

From Luo and Li (2011), Atmospheric Environment

• Idealised CFD study using Reynolds Averaged Navier Stokes equations (RANS)

• Turbulence model: RNG k-ε with standard law of the wall


37

Cooling mountains, warm buildings

• Varying heights of buildings changes the flow, and hence temperature structure

Cooling mountains, warm buildings

• Varying heights of buildings changes the flow, and hence ventilation rate – age of air


38

Sea breeze circulation

Turb ~ 35 C

• Models suggest increase urbanization can lead to stronger sea breeze circulation (Lo et al. 2007, PRD in China)


39

Sea breeze front• Head of sea breeze causes:

‐ uplift of air

‐ enhanced turbulent mixing

• After passage of sea breeze:

‐ shallow, cold layer topped by warmer air

UK Met Office UKV model, 25th July 19:00 UTC: Sea breeze front


40

© Crown copyright Met Office

Thanks to Humphrey Lean, UK Met Office

London, 25th July 2012

• Met Office Unified Model (weather forecast model)

• 80 x 80 km domain • Resolution 100 m• Built fraction• Slab scheme• 140 vertical levels• Turbulence: 3D

Smagorinsky

© Crown copyright Met Office

Thanks to Humphrey Lean, UK Met Office

25th July 2012

Vertical velocity at293m

Unified Model, 100m grid resolution


41

FGAM Doppler Lidar data from the 25th July 2012

Turbulent mixing – associated with passage of sea breeze front?

Ozone obs:150 m,ground

Worst ozone episode since summer 2006

Thanks to Dr James Lee, University of York

High ozone at the ground overnight! Downward mixing by sea breeze??


42

Thermally driven flowsBoundary layer convective mixing entrains air from above BL, and causes a diurnal cycle in wind speed/direction

Flow can be channelled along rivers, with strength modified by CBL mixing

Night-time cooling leads to slope flows…even for shallow slopes!

Tall buildings have their own convective plumes – like a person

Sea breezes interact with the UBL, and can have both negative/positive impacts on air quality


43

Conclusions and challenges• Ventilation of cities occurs at a range of scales

• Some flow patterns can be influenced by good urban design – some

natural processes can be “exploited” ( ~ permaculture)

• Flow is highly non-linear! Details can matter…

• Flow is tricky to measure – observations alone do not give full

picture! Need various methodologies: fullscale, windtunnel/physical

modelling, CFD, mesoscale modelling

• Flow measurement should occur simultaneously at multiple scales –

reference measurement for “city” scale flow, local measurement for

“building” scale flow

• What are the “urban archetypes” needed for more sustainable cities?

Morphology, surface heating/cooling, layout

(more) sustainable city?

water fire

wind earth

ventilation processes from building to city scale

Documents

Transcript of ventilation processes from building to city scale