W hen not to w orry about fugitive PM 1 0 Short tem construction - - PDF document

w hen not to w orry about fugitive
SMART_READER_LITE
LIVE PREVIEW

W hen not to w orry about fugitive PM 1 0 Short tem construction - - PDF document

Fugitive PM 1 0 John W atterson John W atterson ( w ith thanks to John Abbott) ( w ith thanks to John Abbott) W hats in this presentation Guidance Monitoring Modelling Case studies of modelling 1 W hen not to w orry


slide-1
SLIDE 1

1

Fugitive PM 1 0

John W atterson John W atterson ( w ith thanks to John Abbott) ( w ith thanks to John Abbott) W hat’s in this presentation

  • Guidance
  • Monitoring
  • Modelling
  • Case studies of modelling
slide-2
SLIDE 2

2 W hen not to w orry about fugitive PM 1 0

  • “Short tem construction works do not normally

need to be considered for the purposes of Review and Assessment”

  • LAQM.TG4(00), Section 8.36, Page 115

W hich m onitor to use

  • TEOM (Tapered Element Oscillating Microbalance)
  • BAM (Beta Attenuation Monitor)
  • Gravimetric
  • All used in the national networks
  • Partisol (not recommended)
slide-3
SLIDE 3

3 Monitoring at Stage 3

  • LAQM.TG4(00) defines the relevant

criteria and sampling periods

– Ideally monitor for 1 year with 90%

data capture

– minimum 6 months is being advised

(summer/ winter) if comparing with national network sites

– 3 months will often do (but consider

wind direction)

W hen is m onitoring not required?

  • If future predicted concentrations well

below objective, LAs can conclude

– objective unlikely to be exceeded – even if model accuracy only within 50%

  • Recommendation in LAQM.TG4(00) to

monitor in mixed urban areas to validate model

  • At Stage 3 LAs need confidence in their

results

slide-4
SLIDE 4

4 Can you use som eone else’s m onitoring data?

  • Can you use monitoring data from

neighbouring areas / sources to validate your model?

  • For emissions from traffic, yes, but

– area that has been modelled must be ‘identical’ to the one under investigation – (interpret as very similar in nature!)

  • For point sources

– sources need to be similar in nature

For fugitive PM 1 0

  • For fugitive sources

– each source unique, and local meteorological effects likely to be importan so can’t really apply this approach – you’ve probably got to do your own monitoring

slide-5
SLIDE 5

5 QA/ QC of data - it’s essential!

  • Is it necessary to bother with detailed

QA/ QC when it’s

– time consuming – costly – and the data ‘seems reasonable’

  • Yes, because the data cannot be used to

validate the model without QA/ QC

  • An examples of why QA/ QC critical to

follows ...

Data w ithout QC - m onitoring of fugitive PM 1 0

PM10 concentrations (µ µ µ µg m-3 TEOM), without any QC, from fugitive emissions from a quarry

  • 1000
  • 750
  • 500
  • 250

250 500 Mar-99 Jun-99 Sep-99 Dec-99 Mar-00 Jun-00

Time PM10 concentration (µ µ µ µg m-3)

slide-6
SLIDE 6

6 Monitoring for fugitive PM 1 0

Prevailing wind direction Quarry Housing (nearest receptor)

400 m Met. Working face - area of maximum activity PM10 PM10 PM10 sampler Met. meteorological station

Monitoring for fugitive PM 1 0

  • Ideally site PM10 samplers upwind and

downwind close to nearest receptors – but in reality, one sampler upwind

  • Meteorological data is important

– perhaps use data recorded at a nearby site?

  • PM10 filters can be retained for microscopic

analysis

  • Sensitive receptors all > 400 m from active

face? Probably no monitoring needed.

slide-7
SLIDE 7

7 W here can you turn to for m ore help?

  • MONI TORI NG help desk

– Operated by The National Environm ental Technology Centre – Telephone 01 235 46 3 35 6 – Em ail aqm .helpline@aeat.co.uk w eb site w w w .aeat.co.uk/ netcen/ airqual/ w elcom e.htm l

  • MODELLI NG help desk

– Operated by Stanger Science and Environm ent – Telephone 018 1 2 56 4 972 – Em ail m odelhelp@stanger.co.uk – w eb site w w w .stanger.co.uk

  • LOCAL AI R QUALI TY POLLUTANTS' SPECI FI C GUI DANCE

help desk – http:/ / w w w .uw e.ac.uk/ aqm / review – check the FAQ bit

Sum m ary

Correct m onitoring to support assessm ent

!

Ratify the data

"

screen !

"

scale !

"

review Ratify the data

"

screen !

"

scale !

"

review Ensure

"

cost effective m onitoring

"

at the correct locations ( public exposure)

"

  • ver the correct period

"

equipm ent m aintained & calibrated Ensure

"

cost effective m onitoring

"

at the correct locations ( public exposure)

"

  • ver the correct period

"

equipm ent m aintained & calibrated Take tim e to analyse data and present in a useful form Take tim e to analyse data and present in a useful form Use to help validate the m odelling and define areas of exceedence Use to help validate the m odelling and define areas of exceedence

slide-8
SLIDE 8

8

Case studies

Fugitive PM 1 0 em issions John W atterson, Beth Conlan

( w ith thanks Bob W ade, form erly King's Lynn & W est Norfolk Borough Council)

Goods yard - fugitive PM 1 0 issue

  • Overview

– Local Authority wishes to build housing close to a railway – but there are several sources of fugitive PM10 close to the railway line (aggregate delivered by railway) – how close could they build the houses without exceeding the most stringent PM10

  • bjective?
slide-9
SLIDE 9

9 Main sources of em ission

  • Main sources of emission from

aggregate handling operations are

– Loading of aggregate onto storage piles (batch or continuous drop operations); – Equipment traffic in storage area; – Wind erosion of pile surfaces and ground areas around piles; – Loadout of aggregate for shipment or for return to the process stream (batch or continuous drop operations

28500 28550 28600 28650 28700 28750 28800 28850 28900 Easting, m 38600 38650 38700 38750 38800 38850 38900 38950 39000 39050 39100 Northing, m

Site layout & m odelled 9 0 th percentile 2 4 hour average PM 1 0 concentrations, µ µ µ µg m -3

Wet batching plant Aggregate pile Aggregate pile ‘Topmix’ dry batching plant Roadstone coating plant Three sided aggregate storage Railway lines Bottom drop feed hopper Ideal monitor location ACTUAL monitor location

Proposed housing

slide-10
SLIDE 10

10

  • The quantity of dust emitted from

vehicle traffic on a paved road ...

  • The quantity of particulate emissions

from batch or continuous drop

  • perations ...
  • The emissions generated by wind

erosion of stockpiles ...

  • Hourly meteorological data was needed

(March-May 2000)

  • (See me after if you want the detailed

methodology and equations!)

Predictive em ission factor equations for each type of source taken from US EPA AP4 2 em issions database Selection of m onitoring site

  • This was a problem
  • The ‘ideal’ location had no power supply
  • So another location was selected -

location of monitor limited by where there was power

  • Welcome to the ‘real world’!
slide-11
SLIDE 11

11 Receptor locations

  • The dispersion model was used to

calculate the contribution from each modelled source

– to concentrations at receptor locations spaced on a regular grid – extending approximately 500 m from the goods yards – with grid nodes at 50 m intervals.

Model validation for PM 1 0

  • The monitoring data obtained at the

Goods Yard was limited by duration and by the location of the monitoring station

  • Hourly average modelled contributions

from the handling operations compared with the excess concentration at the monitoring station

  • The ‘excess concentration’ was

calculated as the measured concentration at Crawley Goods Yard minus the measured concentration at the Rochester automatic network site (nearest background site)

slide-12
SLIDE 12

12 Data selection criteria for m odel validation

  • The wind was from the sector between 300 and

30 degrees

  • Meteorological data was available from Gatwick

Airport

  • The Crawley Goods Yard monitoring site was
  • perating
  • The Rochester monitoring site was operating
  • There was a train delivering material to the

Aggregates on that day

  • The m odelled concentration was not zero
  • The m easured concentration at Goods Yard was

greater than that at the Rochester site

Model perform ance

  • Measured ‘excess concentration’ plotted

against modelled fugitive PM10 prediction

  • Model overestimating PM10

concentrations by several µgs

  • Considerable uncertainty in the model

predictions as uncertainties inherent in model predictions of emissions and pollutant dispersion and the short period of the model run

slide-13
SLIDE 13

13 W hich concentration to use as exceedence lim it?

  • Uncertainty (one SD) resulting from the

short period of the model run (from period to annual mean monitoring data comparison) was estimated to be 10 µg m -3 (gravimetric)

  • The model validation indicates that the

model overestimates the contribution from the handling operations

  • It is thus likely that the 40 µgm -3 and 60 µg

m -3 isopleths provide reasonable upper and lower bounds on the expected area of exceedence of the 24 hour PM10 objective

Grain handling at a dockyard - fugitive PM 1 0

  • Dust nuisance complaints from grain

handling operations

  • TEOM monitoring for PM10 at a location

nearby suggests exceedences of current 24 hour objective

  • Modelling suggests exceedences of the

24 hour PM10 objective

slide-14
SLIDE 14

14 Data needed for m odelling of fugitive PM 1 0

  • Shipping movements and cargo handled
  • Air quality monitoring

– continuous TEOM to validate the results of the model against

  • Appropriate dispersion model
  • Meteorological data

Monitoring location

slide-15
SLIDE 15

15 Unloading grain from lorry prior to ship loading Grain loading / unloading

slide-16
SLIDE 16

16 Em issions from loading lorries and stockpiles

Response of PM 1 0 concentrations to ship loading activity

Daily average concentrations of PM10 recorded at the South Quay monitoring site 50 100 150 200 250 300 350 400 450 Jan-01 Feb-01 Mar-01 Apr-01 May-01 Jun-01 Jul-01

Date

Concentration of PM10 (µ µ µ µg/m3 gravimetric) South Quay loading Concentration PM10 (µg m-3 grav.) 24 hour objective

slide-17
SLIDE 17

17 Modelling approach

  • Identify suitable emission factors of PM10

for the type of grain loading activity that

  • ccurs at the dock area (US EPA AP 42

database)

  • Compile the detailed activity data available

(periods when commodity handling took place).

  • Estimate emissions of PM10 on an hourly

basis throughout a one year period from all the main areas where grain handling took

  • place. These estimates were derived from

the emission factors and the activity data.

Modelling approach

  • Estimate the 90th percentile of 24 hour

average PM10 concentrations at the location of the nearest PM10 continuous monitoring station. A dispersion model, ADMS V3, and locally measured wind speed and direction data were used to produce these estimates.

  • Calibrate the results of the PM10

modelling using the PM10 concentrations recorded at the PM10 monitoring site.

  • Finally, estimate 90th percentile of

24 hour average PM10 concentrations in an area 600 m by 600 m downwind of the dock areas.

slide-18
SLIDE 18

18

Modelled concentrations of 9 0 th percentile of daily PM 1 0 in 2 0 0 4

Suggested Air Quality Managem ent Area starting betw een the 3 6 µ µ µ µg m -3 and 4 4 µ µ µ µg m -3 contours

Action Plan possibilities

  • Close dow n the grain handling operations
  • Move the operations to an area that w ould

not lead to exceedences at relevant receptors

  • Restrict loading/ unloading operations so

exceedences of the daily PM 1 0 objective unlikely

  • Com plete a process audit to identify
  • perations responsible for m ajority of PM 1 0

em issions

  • Fit abatem ent to control em issions -

cost/ benefit for each of the abatem ent

  • ptions