Motor vehicle sources of particulate matter: Source trends ...€¦ · Receptor modelling source...

GNS Science

Motor vehicle sources of particulate matter: Source trends and related metrics

Perry [email protected]

Transport Knowledge Hub Environment - Emissions Group13 November 2019

mailto:[email protected]

GNS Science

ContextGNS Science has been researching the composition and sources of air particulate matter in NZ (and overseas) for over 20 years.Drivers for this research include:• Understanding human health effects (particle size and

composition)• Air quality management (sources and source

contributions to total PM)• Changes over time (trends and step changes)

― policy evaluation― effectiveness of regulation― impact of technology

GNS Science

Air particulate matter composition and derivation of sources

• A complex mix of elements and compounds from multiple emission sources and atmospheric chemistry (gas↔particle)

• Compositional analysis by nuclear analytical techniques (IBA or XRF) for elements Na to U, black carbon (BC) by light reflectance

• Each source or source type of particulate matter has a distinctive particle size range and chemical composition

• Multivariate covariance analysis and other data analytics across multiple samples to identify sources

Auckland CouncilSea salt Soil Pollen Motor vehicles Wood burner

GNS Science

Auckland PM Speciation monitoring sites

• BC data from 1997 onwards

• Auckland dataset continuous since 2004

Site PM size fraction

Number of filter

samples Sample period

Takapuna PM2.5 1164 November 2006 – June 2016

PM10 1328 December 2005 - December 2018

Queen Street PM2.5 1127 December 2005 - November 2015

PM10 3515 December 2005 - December 2018

Khyber Pass PM2.5 1072 December 2005 - April 2015

PM10 1039 December 2005 – April 2015

Penrose PM2.5 1044 January 2006 - June 2016

PM10 2061 May 2006 - June 2016

Henderson PM10 1250 August 2006 – December 2018

GNS Science

Auckland PM Speciation monitoring sites

All AC sites are either next to motorways:• Penrose, Khyber Pass Rd, TakapunaOr arterial routes:• HendersonOr both:• Khyber Pass Rd, TakapunaOr in a street canyon• Queen Street• The impact of motor vehicle emissions on AC PM

concentrations and composition is significant• The flipside is that we can extract a reasonable amount of

information on what is going on in the (Auckland) motor vehicle fleet

GNS Science

Auckland Council

Biomass combustion

9%

Diesel vehicles

39%

Petrol vehicles

3%

Secondary sulphate

11%

Ship emissions

5%

Marine aerosol

24%

Soil2%

Construction7%

Queen Street average PM2.5 = 9.3 µg m-3

0

5

10

15

20

Dec-05 Jun-06 Dec-06 Jun-07 Dec-07 Jun-08 Dec-08 Jun-09 Dec-09 Jun-10 Dec-10 Jun-11 Dec-11 Jun-12 Dec-12 Jun-13 Dec-13

BC

con

cent

ratio

n (µ

g m

-3) Biomass burning

0

5

10

15

20


BC

con

cent

ratio

n (µ

g m

-3) Diesel vehicles

0

5

10

15

20


BC

con

cent

ratio

n (µ

g m

-3) Petrol vehicles

0

5

10

15

20


BC

con

cent

ratio

n (µ

g m

-3) Secondary sulphate

0

5

10

15

20


BC

con

cent

ratio

n (µ

g m

-3) Ship emissions

0

5

10

15

20


BC

con

cent

ratio

n (µ

g m

-3) Marine aerosol

0

5

10

15

20


BC

con

cent

ratio

n (µ

g m

-3) Soil

0

5

10

15

20


BC

con

cent

ratio

n (µ

g m

-3) Construction

Source apportionment

0

10

20

30

40

Dec-05 Dec-06 Dec-07 Dec-08 Dec-09 Dec-10 Dec-11 Dec-12 Dec-13 Dec-14 Dec-15

PM2.

5co

ncen

tratio

n (µ

g m

-3)

• PM compositional analysis used to identify sources

Queen Street PM2.5 example

GNS Science

Sources of particulate matter across Auckland

• Data for common sources aggregated for all sites• Anthropogenic emissions dominate PM2.5 (≈ 70 %)

– Vehicles, biomass combustion, soil, some secondary sulphate• Natural sources significant for PM10 (≈ 50 %)

PM2.5 PM10

Biomass combustion

18%

Diesel vehicles

36%Petrol

vehicles7%

Secondary sulphate

12%

Marine aerosol

19%

Soil2.7%

Other5%

PM2.5 = 7.5 µg m-3

Biomass combustion

11%

Diesel vehicles

19%

Petrol vehicles

2%

Secondary sulphate

8%

Marine aerosol

45%

Soil4.6%

Other11%

PM10 = 16.4 µg m-3

GNS Science

Trends in particulate matter across Auckland• Data aggregated for all PM speciation sites• Downward trend observed for PM2.5 and PM10 (99.9% CI)• PM2.5 apparently responsible for decrease in PM10

• Understanding ‘why’ important for AQ management

PM2.5 PM10

GNS Science

Motor vehicle source activity

Auckland Council

• Difference in weekday/weekend PM concentrations― This and crustal matter are the only sources that show this dichotomy― Less HCV (diesel) activity during weekends

• Mondays affected by public holidays (≈ 15%)• Difference between PM2.5 and PM10 contributions is road dust

component

PM2.5 PM10

Johnstone Hills Tunnel

Davy PK, Trompetter WJ, Markwitz A. 2011. Concentration, composition and sources of particulate matter in the Johnstone Hills Tunnel, Auckland. 64 p. (GNS Science consultancy report; 2010/296)

Light duty diesel

vehicles27%

Heavy commercial diesel vehicles

67%(10% of total

vehicle counts)

Petrol vehicles

6%

Vehicle PM by time of day

Takapuna vehicle PM

Diesel vehicles

90%

Petrol vehicles

10%

GNS Science

AC Transport PM source trends

• Diesel vehicle related PM decreasing (Technology impact)– Engine design/emission

improvements50 % of PM2.5 trend ↓

• Primary tailpipe component is a reduction in BC emissions (representing ultra-fines)So it’s a good news story!

• Petrol vehicle PM no change although vehicle numbers increasing (Technology impact)– Fuel efficiency gainsNo impact on PM2.5 trend

2008 dip then rise

GNS Science

BC concentrations by source in Auckland• Influence of roadside sites on average BC source

concentrations (2006 – 2018) evident• Shift in bus route resulted in BC step change at Queen Street

– Just shifted the impact to another street canyon Localized effects of policy changes

0

1000

2000

3000

4000

Queen Street Khyber PassRoad

Penrose Takapuna Henderson

Blac

k car

bon

ng m

-3

Biomass burning

Diesel vehicles Bus route change

Queen Street BC

GNS Science

1 Jan 2008Petrol 50 ppm

1 May 2007Petrol 150 ppmDiesel 50 ppm

1 Jan 2009Diesel 10 ppm

• Reduction in sulphate concentrations (2006 - 2010) as a consequence of removal of S in fuels‒ Petroleum Products Specification Regulations 2002, 2007

Transport source trends

New urban ‘steady-state’Anthropogenic + natural SO4

2-

GNS Science

Transport source trends Road dust – a mixed bag• Contributions to PM are location

dependent– Proximity to road– Vehicle size, number, speed– Road surface type

• Multiple components– Brake wear (Fe, Cu, Zn, Ba, Sb, Ni….)– Tyre wear (Zn, S, BC)– Road surface wear (Al, Si, Ca…)– Deposited material from other sources

• Electric vehicles will not reduce road dust component

Khyber Pass Road increasing

Queen Street decreasing

GNS Science

Impact zones of PM from motor vehicles –Khyber Pass road example• Highest concentrations aligned with centreline of motorway

(South East ‘fetch’ 20km+)

GNS Science

Transport source trends - PM10 brake dust copper• Reflect road dust component and traffic volumes• Also implications for deposition and water runoff• Impact may change with brake pad composition (e.g. asbestos)

Khyber Pass Road increasing

Takapuna increasing

PenroseNo change

Henderson increasing

Queen Street decreasing

GNS Science

Transport activity trends (Fuel consumption)Domestic land transport data

• Diesel fuel consumption rising– Engine improvements

offset PM emissions from more fuel use

– (check whether this includes off-road consumption?)

• Petrol consumption flat– vehicle numbers

increasing (fuel efficiency gains)

• HCV freight (tonne-km) increasing (note 15% dip at 2008-2009)

0

50

100

150

200

250

300

2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016

Gros

s PJ

Observed Oil Products Consumption

LPG

Other PetroleumProductsAviation Fuels

Fuel Oil

Diesel

Petrol

Source: Ministry of Business, Innovation and Employment (2018)

Legend

16000

18000

20000

22000

24000

26000

60

80

100

120

140

160

180

2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016

Tonn

e -k

m

Gros

s PJ


Diesel

Freight tonnage


GNS Science

Trends in PM contributions and traffic volumes

Auckland Council

• Emissions improvements offset by local traffic volume increase

• Macro-economic effect: Global Financial Crisis

70,000

90,000

110,000

130,000

150,000

170,000

190,000

0

1

2

3

4

5

6

2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016

Traf

fic v

olum

e (A

ADT)

Conc

entra

tion

(µg

m-3

)

Takapuna: PM2.5 and PM10 from vehicles and traffic volumes

Traffic volume PM2.5 PM10 Linear (Traffic volume)

100,000

120,000

140,000

160,000

180,000

200,000

220,000

0

1

2

3

4

5

6

2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016

Traf

fic v

olum

e (A

ADT)

Conc

entra

tion

(µg

m-3

)

Penrose: PM2.5 and PM10 from vehicles and traffic volumes

Traffic volume PM2.5 PM10 Linear (Traffic volume)

Road dust

Road dust

Penrose Takapuna

Motor vehicle PM2.5 ↓ ―

Traffic volume ― ↑

Motor vehicle PM10 ↓ ↑

16000

18000

20000

22000

24000

26000

60

80

100

120

140

160

180

2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016

Tonn

e -k

m

Gros

s PJ


DieselFreight tonnage


GFCGFC

GFC

Takapuna

Penrose

GNS Science

Sector Air Quality Impacts – Transport metrics

Auckland Council

Fuel economy/efficiency (L/100km) – Success!

Fleet size and composition (diesel/petrol) - Challenge

NZ new or used import – Success!

Land Transport Rule 2007 Vehicle Exhaust Emissions

Vehicle kilometres travelled - Challenge

Fleet age - Challenge

Fuel consumption -Challenge

DieselPetrol

Motor vehicle tailpipe PM emissions

Success!

Motor vehicle tailpipe gas (CO2, NOx) and road dust emissions

Challenge!but

Success!

GNS Science

Direct PM emissions from ships engines• Represent primary emissions from ships engines – this

would also include acid sulphur species, PAHs etc• Ratio of V/Ni measured at Queen Street site is 3:1 same

as heavy fuel oils

Queen Street (monthly averages)

GNS Science

Direct PM emissions from ships engines

• Receptor modelling source elemental profile reflects the primary combustion product components (BC, S, V, Ni)

• Ratio of V/Ni is 3:1 same as heavy fuel oils

0

1

10

100

1000

H BC Na Mg Al Si S Cl K Ca Ti V Mn Fe Ni Cu Zn Ba

Conc

entr

atio

n ng

m-3

Ship emissions source profile

GNS Science

Queen Street CPF plot shows shipping source arrives at the site from the port area

GNS Science

Trends in Auckland ship emission contributions• Concentrations derived from receptor modelling increasing

(99.9 % CI) - reflects shipping activity at the port• Main impact is on Auckland CBD

Ports of Auckland Annual report 2018

GNS Science

Summary• Engine technology and regulation (fuel improvements,

import emission standards) have had a significant downward effect on PM2.5 …but traffic volume increases may offset this particularly for PM10

• Differential impact on PM by motor vehicle type– Diesel vs petrol

– HCV vs light duty

• Shipping impacts are increasing Future work• Combine PM source apportionment with transport

metrics for robust statistics• Signal processing of PM2.5/PM10 data

Motor vehicle sources of particulate matter: Source trends ...€¦ · Receptor modelling source...

Documents

Transcript of Motor vehicle sources of particulate matter: Source trends ...€¦ · Receptor modelling source...