Appendix 6 Modelling for Atmospheric Concentrations of Formaldehyde

A1. Modelling methodology

In the calculation procedures described, a number of different approaches have been adopted to calculate PECs, depending on the type of source. In each case, the maximum annual average and maximum 24-hour average concentrations have been computed. The conversion 1 ppb = 1.2 µg/m³ for formaldehyde has been used.

• 
  For individual industrial sources, year-long modelling with AUSPLUME version 5.4 (a regulatory model developed by EPA Victoria) (EPA Victoria, 2000) was carried out using a 1997-1998 meteorological data file for Paisley in the western part of Melbourne. This meteorological file was derived using data from The Air Pollution Model (TAPM) modelling of the urban region described in Section A3 below. AUSPLUME is a Gaussian plume dispersion model, which is suitable for predicting ground-level concentrations of pollutants from a variety of sources. In addition to the emission rates (derived from 2001–2002 NPI data), the modelling requires information on the source configuration, for example, whether it is a diffuse area source, a fugitive emission from a building, or a release from a chimney stack. A detailed analysis would require details of the source characteristics for each facility in the NPI database, which is beyond the scope of this modelling. Instead, estimates of the source configurations were based on the source descriptions provided by NICNAS and summarised in Section 8 of the assessment report. Concentrations were calculated at a distance of 100 m from the source, except in cases where the maximum occurs at a greater distance (e.g. for tall stack releases), in which case the maximum PECs are reported. The distance of 100 m is representative of the distance from the source to the boundary of an industrial site. For near-surface sources, concentrations decrease at greater distances from the boundary. AUSPLUME was run for 1997-1998 for each source and so included the full range of meteorological conditions and stabilities. The results were analysed to derive the maximum 24-hour average concentration and the annual average concentration for each source. Separate calculations were made for the
  

average emitter and the largest emitter. Based on the information available about the sources, in some cases, it was appropriate to use different source configurations for the average and the largest emitters.

• 
  For diffuse urban sources, results were obtained from re-analysis of a detailed urban airshed modelling study of Melbourne undertaken for EPA Victoria by CSIRO in 2001 (Hurley et al., 2001). This study used the most up-to-date spatially-distributed emissions inventory for the region for a large number of pollutants. The modelling was carried out using TAPM. This model was developed at CSIRO Atmospheric Research Division (Hurley, 2002) and consists of prognostic meteorological and air pollution modules that can be run for multiple-nested domains. The meteorological module is an incompressible, non-hydrostatic, primitive equation model for three- dimensional simulations. It predicts the three components of the wind, temperature, humidity, cloud and rainwater, turbulent kinetic energy and eddy dissipation rate, and includes a vegetation/soil scheme at the surface and radiation effects. The model is driven by six-hourly analysis fields (on an approximately 100-km spaced grid) of winds, temperature and specific humidity from the Bureau of Meteorology’s Global Assimilation and Prediction system (GASP). These analyses contain the larger-scale synoptic variability, while TAPM is run for much finer grid spacings and predicts the meteorology at smaller scales. The air pollution module solves prognostic equations for pollutant concentration using predicted wind and turbulence fields from the meteorological module. The modelling used in the original report (Hurley et al., 2001) was carried out with 20 20 20 point nested grids at 30-km, 10-km, 3-km and 1-km horizontal grid spacings for the year July 1997 to June 1998. The re- analysis generated 24-hour averages, which supplemented the results for annual average concentrations for formaldehyde presented in the original report.
  

• 
  For roadway emissions, an idealised large urban freeway was modelled using AUSROADS (EPA Victoria, 2002), with the concentration calculated at distances of 0, 20 and 100 m from the edge of the freeway. The same 1997-1998 meteorological data file was used as for the industrial source modelling. AUSROADS is a line source Gaussian plume dispersion model that predicts the near-road impact of vehicle emissions in relatively uncomplicated terrain (EPA Victoria, 2002).
  

The models do not include effects of secondary formation and destruction of formaldehyde, which can have an impact on PECs. However, the annual average is probably dominated by poor dispersion in winter when the inversion level is low (i.e. a smaller volume of air in which the formaldehyde is mixed) rather than by secondary production in summer when the mixing height is much greater. Furthermore, it is expected that these effects are small compared to the uncertainties in emission rates and source configurations used in the modelling.

A2 Industrial source impact modelling

NICNAS identified seven industrial source categories and used the NPI database to derive emission data for each category (Section 13.1.1). The average and maximum emission rates are summarised in Table A6-1.

Modelling was carried out using the Gaussian plume model AUSPLUME version 5.4. A one-year meteorological file was derived from the TAPM modelling of 1997-1998 (described in Section A3 below) for a site near the Paisley air quality monitoring site in the west of Melbourne.
The following section provides details of the emission rates from each industry category and the source configurations used in the AUSPLUME modelling. The results shown are the predicted maximum annual average and maximum 24-hour average concentrations at a distance of 100 m from the source, except where the maximum occurs at a greater distance (e.g. for tall stack releases), in which case the maximum PECs are reported.
Table A6-1: Average and maximum annual emissions of formaldehyde for each industry category and source configuration used in the AUSPLUME modelling
Type of industry Annual release rate (kg/year) Source details used in modelling for

average industry source
Average Maximum from

an individual facility
^Mining 12 203 401 112 ^{surface and near-surface sources,}

1000 m diameter area source for average emitter

^{Wood & paper 8195} 27 082 ^{releases from process and storage}

areas

30 m stack (50% of emissions) and fugitives at 10 m (50% of emissions)



Electricity supply


Materials manufacture
⁴⁷⁹² 85 614 ^{combustion product released via} stacks

50 m stack for average emitter 200 m stack for largest emitter

³⁶⁶⁴ 35 000 ^{releases from buildings and short} stacks

30 m stack (50% of emissions) and fugitives at 10 m (50%)

Petroleum 3162 8883 refinery combustion released via

stacks

50 m stack

Chemical manufacture
651 6960 most emissions via stacks

30 m stack next to a 20 m high building

Miscellaneous 79 1099 releases from building and process

areas

fugitives released from a 10 m building

A2.1 Mining operations

For the average source with an annual release rate of 12 203 kg/year, a representative source configuration was assumed to be a surface source with a diameter of 1000 m and an initial vertical spread of 10 m. Using AUSPLUME modelling, the annual average PEC at 100 m from the edge of the activity was 1.8 ppb and the maximum 24-hour average was 8.1 ppb. These results are approximately inversely proportional to the diameter of the area source (for a given emission rate).

Given that the main sources of emissions from mining operations are distributed surface sources, the area of emissions is likely to be approximately proportional to the emissions rate, so that PECs from the largest emitter (401 112 kg/year) are expected to be similar to those from the average emitter.

A2.2 Wood and paper product manufacturers

For the AUSPLUME modelling it was assumed that the emissions are split between two points: 50% from a 30 m stack (with a diameter of 2 m, efflux velocity of 10 m/s, and temperature of 25ºC) and 50% as fugitive emissions at a height of 10 m (represented as a volume source released at a height of 10 m and with an initial vertical and horizontal spread (two times the standard deviation) of 10 m). The annual average PEC 100 m from a facility with an average emission rate was 4.8 ppb and the maximum 24-hour average was 36 ppb. The highest estimated PECs from the largest emitter were 16 ppb (annual average) and 119 ppb (maximum 24-hour average). A sensitivity analysis showed that the PECs are much more sensitive to the configuration of the source of the fugitive emissions than the stack emissions. All of the wood and paper product industries in the NPI database are located outside major urban areas. However, given the high PECs, it would be appropriate to verify these predictions by obtaining more information about the source configurations for these industries.

A2.3 Electricity supply

The source configuration for the average emitter was assumed to be a 50 m stack (2 m diameter, 10 m/s efflux velocity, and a temperature of 25ºC). This produced PECs of

0.11 ppb (annual average) and 1.12 ppb (maximum 24-hour average). For the largest emitter, the source was taken to be a 200 m stack (3 m diameter, 20 m/s efflux velocity, 25ºC), which produced similar PECs of 0.10 ppb (annual average) and 0.98 ppb (maximum 24-hour average) due to the greater release height. These PEC estimates are conservative because buoyant plume rise was ignored by setting the efflux temperature to 25ºC. The largest emitter has slightly lower PECs than the average emitter because of the higher release height.

A2.4 Materials manufacture

The source configuration for the average facility is assumed to split between two points: 50% from a 30 m stack (diameter of 2 m, efflux velocity of 10 m/s, and temperature of 25ºC) and 50% as fugitive emissions at a height of 10 m (represented as a volume source released at a height of 10 m and with an initial vertical and horizontal spread (two times the standard deviation) of 10 m). The PECs from the AUSPLUME modelling are 2.1 ppb (annual average) and 16 ppb (maximum 24-hour average).

For the largest emitter (an aluminium refinery) the source is taken to be a 50 m stack (2 m diameter, 10 m/s efflux velocity at 25ºC). This produces PECs of 0.78 ppb (annual average) and 8.2 ppb (maximum 24-hour average). These values are lower than for the average emitter because the emission occurs from a taller stack.

A2.5 Petroleum refining, oil and gas extraction

The point source emissions are assumed to occur from a 50 m stack (2 m diameter, 10 m/s efflux velocity at 25ºC). For the average emitter, AUSPLUME modelling produces PECs of 0.07 ppb (annual average) and 0.74 ppb (maximum 24-hour average), whereas the

largest emitter (8883 kg/year) produces PECs of 0.20 ppb (annual average) and 2.1 ppb (maximum 24-hour average).

A2.6 Chemical industry

As most emissions occur via stacks, the source configuration used for the AUSPLUME modelling was a 30 m stack (2 m diameter, 10 m/s efflux velocity, 25ºC temperature) next to a 20 m high building. Maximum concentrations were found to occur 300 to 500 m from the stack. For the average facility (651 kg/year), the maximum annual average PEC was

5. 
   ppb and the maximum 24-hour average was 0.41 ppb. For the largest formaldehyde manufacturing plant (6960 kg/year), the maximum annual average PEC was 0.57 ppb and the maximum 24-hour average was 4.4 ppb.
   

A2.7 Miscellaneous industries

In the AUSPLUME modelling, the source configuration was taken to be fugitive emissions from a 10 m building (represented as a volume source released at a height of 5 m and with an initial vertical and horizontal spread (two times the standard deviation) of 5 m). For an average emitter, the PECs were 0.14 ppb (annual average) and 1.2 ppb (maximum 24-hour average). For the largest emitter, the PECs were a factor of 14 larger, namely 2.0 ppb (annual average) and 17 ppb (maximum 24-hour average).

A2.8 Summary of point source PECs

Table A6-2 summarises the results predicted environmental concentrations described in the above sections. In each case, the maximum annual average and maximum 24-hour average PECs are listed for the average emitter and for the largest emitter.

The highest PECs occur for the wood and paper industries, which are the second largest emitter after mining operations (Table A6-1). The impact on PECs from the wood and paper industries is greater because the concentrations at the release points is higher than in mining (which is mainly due to vehicles and other surface sources). However, all of the wood and paper product industries in the NPI database are located outside major urban areas, so that they will not impact on formaldehyde concentration in large urban areas. In spite of this, the high PECs indicate that it would be useful to verify these predictions by obtaining more information about the source configurations for these industries to check the modelling assumptions.

A3 Urban impact modelling

Urban levels of formaldehyde due to diffuse urban emissions were determined from a re- analysis of detailed urban airshed modelling of ambient pollutant concentrations in Melbourne undertaken by CSIRO for EPA Victoria (Hurley et al., 2001). The original study, which was part of work for the EPAV Air Quality Improvement Plan, used a comprehensive inventory of emissions from industry, motor vehicles (petrol fuel type), wood heater emissions and biogenic emissions. The modelled year was 1997/98. The re- analysis generated 24-hour averages to supplement the original modelling of annual average concentrations. The results from the modelling with a 3-km grid spacing are listed in Table A6-3. This grid spacing of 3 km was used because it minimises the local impact from some industrial sources and thus provides an estimate of urban concentrations away from significant local sources, such as industry or large roads. The annual average concentration is 1.6 ppb and the maximum 24-hour average is 13 ppb.

Table A6-2: Summary of maximum annual average and maximum 24-hour average predicted environmental concentrations calculated from AUSPLUME modelling for each industry category



Type of industry Maximum Annual Average PECs
Maximum 24-hour Average PECs

	Average emitter	Largest emitter	Average emitter	Largest emitter
Mining	1.8 ppb	≈1.8 ppb	8.1 ppb	≈8.1 ppb
Wood & paper Electricity supply	4.8 ppb 0.11 ppb	16 ppb 0.10 ppb	36 ppb 1.12 ppb	119 ppb 0.98 ppb
Materials manufacture Petroleum Chemical manufacture	2.1 ppb 0.07 ppb 0.05 ppb	0.78 ppb 0.20 ppb 0.57 ppb	16 ppb 0.74 ppb 0.41 ppb	8.2 ppb 2.1 ppb 4.4 ppb

Miscellaneous 0.14 ppb 2.0 ppb 1.2 ppb 17 ppb


Table A6-3: Maximum formaldehyde concentrations in the Melbourne urban region from TAPM modelling for the year July 1997 to June 1998 (3-km grid spacing)

Averaging time Maximum formaldehyde
70^th percentile

concentration

Annual average	1.6 ppb	-
24-hr average	13 ppb	2.2 ppb

When determining the impact of an industrial source located in an urban area, it is common practice to add the maximum PEC for the industrial source to a typical urban background concentration, represented by the 70^th percentile (EPA Victoria, 1985), rather than the maximum urban background, which is unlikely to occur at the same time as the maximum source impact. Figure A6-1 shows the cumulative probability distribution of the 24-hour averages with the 70^th percentile which equals to 2.2 ppb.

A4 Near road impact modelling

Maximum formaldehyde concentrations due to roadway emissions were determined from modelling of emissions from a 6-lane dual carriageway freeway. Modelling was carried out using AUSROADS, which is a Gaussian dispersion model based on the Caline-4 model, with a user-friendly interface developed by EPA Victoria (EPA Victoria, 2002). The modelled roadway geometry was a straight section 3 km in length with 3 lanes in each direction, representative of a large urban freeway. Each lane was 4 m wide and there was a separation of 8 m between the carriageways. The total daily flow rate was modelled to be 150 000 cars per day, evenly divided between each of the 6 lanes with the diurnal distribution shown in Figure A6-2. This diurnal distribution was based on weekday flows

in the Sydney M5 East tunnel and Melbourne's CityLink tunnel, scaled up to 25 000 vehicles per lane per day, which was considered to be typical of city freeway flows.

Figure A6-1: Annual cumulative probability distribution of 24-hour average formaldehyde concentrations in the Melbourne urban region from re- analysis of TAPM modelling for the year July 1997 to June 1998 (3-km grid spacing)

0.999

0.998
0.995

0.99

0.98
Max = 13 ppb

Cumulative probability

0.95
0.9
0.8
0.7

0.6

0.5

0.4

0.3
0.2

0.1

0 2 4 6 8 10

24-hour average formaldehyde concentration (ppb)

Figure A6-2: Assumed diurnal variation in traffic flow on each lane of the modelled roadway for a total daily flow of 25 000 vehicles per lane.
2000



Vehicles per lane per hour

1500

1000

500

0
0 6 12 18 24

Hour of Day

The fleet average emission factor for formaldehyde was taken to be 20 mg/km, as reported in a Melbourne study by EPA Victoria (1999c). This compares with a recent value of 13.7 mg/km reported for measurements on Melbourne’s CityLink (Tran et al., 2003). However, the lower value for CityLink traffic probably reflects the higher proportion of newer cars (with reduced emissions) than would be found in a city-wide average.
The meteorological data file for AUSROADS was the same as that used in the AUSPLUME modelling discussed in Section A2 above, i.e. for 1997-1998 and representative of the western region of Melbourne. To remove any influence of a predominant wind direction from the results, modelling was carried out with the roadway aligned north-south and then east-west. Less than 10% difference was found between the maximum predicted concentrations for these two orientations. The results at three distances from the edge of the roadway, listed in Table A6-4, show a rapid decrease with increasing distance from the roadway.
Table A6-4: PECs for typical large urban freeway (150 000 cars per day) modelled using AUSROADS


Location Maximum annual average PEC

Maximum 24-hour average PEC

At edge of freeway	0.77 ppb	2.3 ppb
20 m from edge of freeway	0.37 ppb	1.06 ppb

100 m from edge of freeway 0.15 ppb 0.50 ppb

Directory: data -> assets -> word doc -> 0006
word doc ->  Commonwealth of Australia 2002
word doc -> Commonwealth of Australia 2000
word doc -> Priority Existing Chemical
word doc -> Rail safety news issue 6 – October 2011
word doc -> Review of Multiple Chemical Sensitivity: Identifying
0006 -> Board Endorsed October 2009 amended 13-09-11 with extra units
0006 -> Rt Hon Andrew Lansley cbe mp secretary of State for Health
0006 -> Environmental Approvals Application Form Use this form

Download 5.77 Mb.

Share with your friends: