Correcting TEOM Measurements using the KCL Volatile

Correction ModelDavid Green, Gary Fuller

& Timothy Baker

2 Presented by David Green

Contents

• Model Development Context• UK PM Monitoring Problem• Model derivation• Model testing• Proposed monitoring strategy for

the UK

3 Presented by David Green

Model Development Context

• January LAQN Seminar – London FDMS• Defra funded work early 2007

–Grounded in statistics used to demonstrate equivalence

– UK wide applicability

• Named it: “Volatile Correction Model”• Future steps

4 Presented by David Green

PM10 Monitoring Problem• Complex pollutant

– Different components

– Constantly changing

• Reference Method– Labour intensive

– Poor time resolution

– Slow data dissemination

Reference

MethodTEOM

FDMS

• TEOM– Heats sample inlet to 50°C to

eliminate water but looses volatile PM

• FDMS– Uses a diffusion dryer to eliminate

water and retain a sample temperature of 30°C

– Equivalent to Reference

• UK Monitoring Networks– Predominately TEOM

– Need to use equivalent methods for reporting to EU

BAM

5 Presented by David Green

PM10 Measurement Time Series

0

10

20

30

40

50

60

70

28/11/04 28/01/05 28/03/05 28/05/05

Da

ily

Me

an

Co

nc

en

tra

tio

n (

µg

m-3

)KFGTEOM

6 Presented by David Green

PM10 Measurement Time Series

0

10

20

30

40

50

60

70

28/11/04 28/01/05 28/03/05 28/05/05

Da

ily

Me

an

Co

nc

en

tra

tio

n (

µg

m-3

)KFGTEOMFDMS

7 Presented by David Green

PM10 Measurement Correlation

0

10

20

30

40

50

60

0 10 20 30 40 50 60

Reference Method (µg m-3)

Ca

nd

ida

te M

eth

od

(µ

g m

-3)

TEOM

FDMS

8 Presented by David Green

How do we achieve an Equivalent Network?

• Upgrade TEOM to FDMS–Expensive (capital)

–Retains some continuity of measurement

• Change monitoring equipment–Gravimetric

– (capital and revenue)

– Loose continuity of measurement

–Delay in reporting time

–BAM–Expensive (capital)

– Loose continuity of measurement

• A ‘Third Way’?–Using FDMS measurements of volatile PM to

correct TEOM measurements

9 Presented by David Green

KCL Volatile Correction Model

• Provides a daily, site specific correction factor for TEOM measurements

• Correction based on FDMS purge measurement made some distance away

• Results in reference equivalent daily mean concentration within the 25% expanded uncertainty specified by the AQ Directive

10 Presented by David Green

Model Derivation

11 Presented by David Green

FDMS – Filter Dynamics Measurement System

• 2 measurement modes:–Base (analogous to

standard TEOM)

–Purge, which measures mass lost from the filter when particle free air is passing through it

• FDMS Mass = Base - Purge

12 Presented by David Green

Model Derivation

13 Presented by David Green

TEOM and FDMS Monitoring

East Kilbride

Birmingham

Bristol

Teddington

14 Presented by David Green

TEOM – Base vs. FDMS Purge

-25

-20

-15

-10

-5

0

5

23/12/2004 20/01/2005 17/02/2005 17/03/2005 14/04/2005 12/05/2005 09/06/2005

Co

nce

ntr

atio

n (

µg

m-3

)

Birmingham TEOM-Base Birmingham Purge

15 Presented by David Green

TEOM – FDMS Base vs. FDMS Purge

-25

-20

-15

-10

-5

0

5

-25 -15 -5 5

TEOM - FDMS Base (µg m-3)

FD

MS

Pu

rge

(µg

m-3

)

16 Presented by David Green

PM10 Measurements

0

10

20

30

40

50

60

0 10 20 30 40 50 60

Reference Method (µg m-3)

Ca

nd

ida

te M

eth

od

(µ

g m

-3)

TEOM

FDMS

17 Presented by David Green

PM10 Measurements

0

10

20

30

40

50

60

0 10 20 30 40 50 60

Reference Method (µg m-3)

Ca

nd

ida

te M

eth

od

(µ

g m

-3)

TEOM

FDMS

Model

18 Presented by David Green

Model Derivation

19 Presented by David Green

Uniform FDMS Purge Concentrations

-25

-20

-15

-10

-5

0

5

23/12/2004 20/01/2005 17/02/2005 17/03/2005 14/04/2005 12/05/2005 09/06/2005

Co

nce

ntr

atio

n (

µg

m-3

)

Bexley Purge North Kensington Purge Marylebone Purge

20 Presented by David Green

Equivalence Testing

• Experiment 1 – test the model at the equivalence programme sites excluding regional aspects

• Experiment 2 - test the model at the equivalence programme sites including regional aspects

Criteria

n ≥ 40

n ≥ 50 % of limit value ≥ 25%

Between reference sampler uncertainty

≤ 2 µg m-3

Between candidate sampler uncertainty

≤ 3 µg m-3

Expanded Uncertainty ≥ 25%

21 Presented by David Green

Experiment 2

22 Presented by David Green

Expanded Uncertainty

0

10

20

30

40

50

60

70

80

90

0 200 400 600

Distance (km)

Exp

and

ed U

nce

rtai

nty

(%

)

Model

FDMS

23 Presented by David Green

Monitoring Strategy based on the VCM Model

24 Presented by David Green

Conclusion

• Model provides a daily, site specific correction factor for TEOM measurements to provide a reference equivalent measurement:

Reference Equivalent PM10 = TEOM – 1.87 FDMS purge

• Works up to a distance of 200 km• Allows smaller number of FDMS instruments

to correct larger network of TEOMs–Financial and data continuity implications

• Further work…

25 Presented by David Green

Further Work

• Physical and chemical basis for model–Concentrations on TEOM and FDMS filters

–Ammonium nitrate

–Volatile organic compounds

–Collocated measurements–Ammonium nitrate

–Volatile organic compounds

–Water

• Extend to hourly public dissemination• Provide method for local authorities to use the

model• Extend to PM2.5

26 Presented by David Green

Acknowledgements

• Defra• London Borough of Bexley• London Borough of Greenwich• London Borough of Ealing• City of Westminster• Royal Borough of Kensington and

Chelsea• Air Monitors

27 Presented by David Green

Thank you for you attention!