This work is copyright. Apart from any use as permitted under the Copyright Act 1968, no part may be reproduced by any process without prior written permission from the Commonwealth
Download 5.77 Mb.
|
- Navigate this page:
- Report No 79365
- MDF PLANT FORMALDEHYDE AIR DISPERSION
- INTRODUCTION
- ATTACHMENT 1 AUSTRALIAN WOOD PANELS ASSOCIATION
- ATTACHMENT 2 AUSTRALIAN WOOD PANELS ASSOCIATION
Appendix 17
AUSTRALIAN WOOD PANELS ASSOCIATIONMDF PLANT FORMALDEHYDE AIR DISPERSION Report No 79365May 2006 EML AIR PTY LTD ABN 98 006 878 342 417-431 Canterbury Road Surrey Hills VIC 3127 Telephone (03) 9836 1999 Facsimile (03) 9836 0517
EXECUTIVE SUMMARYAustralian Wood Panels Association (AWPA) requested that the dispersion of formaldehyde from an MDF Plant emitting 27,000 kilograms per year be compared for two (2) scenarios. The first air dispersion modelling scenario attempts to replicate the model used by CSIRO Atmospheric Research “Formaldehyde Air Quality Assessment” Report C/0928 for NICNAS May 2004. This scenario used the annual national pollutant inventory data for a large wood products plant from year 2000. The formaldehyde emissions to air were assumed by CSIRO to be split, 50 percent each between stack and fugitive sources, for an assumed flat ground site location. The second scenario, described in the following report, considers the same quantity of annual formaldehyde emissions being emitted from both stack and fugitive sources but based on a more representative site source emission profile. The emission profile used in the model has been distributed to the various sources as follows:
The peak predicted 24-hour average and annual average ground level concentrations from both air dispersion model scenarios are compared and discussed. Using a typical plant layout and test emission profile for the stack sources has predicted formaldehyde ground level concentrations, at the same locations, which are an order of magnitude lower than the CSIRO predicted concentrations.
Australian Wood Panels Association requested that the dispersion of formaldehyde from an MDF Plant emitting 27,000 kgs p.a. be compared in two (2) scenarios. The first would review the CSIRO Atmospheric Research “Formaldehyde Air Quality Assessment” Report C/0928 for National Industrial Chemicals Notification and Assessment Scheme (NICNAS) May 2004 and the second as typical plant layout and source emission profile. The Ausplume Guassian plume dispersion model (EPAV Version 6) has been used for these predictions of ground level concentration (g.l.c.) 24-hour average and annual averages.
The draft report by NICNAS “Formaldehyde: Priority Existing Chemical Assessment” in Appendix 6 “Modelling of Atmospheric Concentrations of Formaldehyde” Section A1, Modelling Methodology, acknowledges a potential significant source for error in the predicted environmental concentrations (PEC`s):
Furthermore the National Pollutant Inventory (NPI) Timber and Wood Product Manufacturing Emission Estimation Factors for the particle board dryers and press emissions are coded as D (Below average) and E (Poor). Where Australian facilities have not had available extensive emission tests conducted then considerable variances can arise from actual emissions should the modelling be conducted on data from a facility that has been mostly reliant on using the NPI factors. The Appendix 6 Section A.2.2, ‘Wood and Paper Product Manufacturers’ has further commented that:
Table A.6-2 Summary of predicted maximum annual and 24-hour average PEC`s concludes that for Wood and Paper Industries that average emitters may have concentrations 100 metres from the plant of 4.8 ppb and 36 ppb respectively. These average PEC`s are consistent with the requirement of EPA`s in Australia that new or altered facilities had to meet dispersed formaldehyde ground level concentrations of 50 ppb (3-minute average) up until the late 1990`s and since then, 33 ppb (3-minute average). However the predicted maximum annual and 24-hour average PEC`s concludes that for Wood and Paper Industries largest emitters the concentrations at 100 metres could be 16 ppb to 119 ppb respectively. The highest predicted 24-hour average is significantly at variance with the following:
These three aspects are indicative that the modelling by CSIRO for NICNAS has generally over-predicted ground level concentration of formaldehyde at the boundary of wood product industries, particularly in the case of largest emitters. The actual annual and 24-hour average concentrations are more likely to be similar to the PEC`s predicted for the other 6 categories of large emitter industries listed in the CSIRO report Table A6-2.
The AWPA requested that the model used by CSIRO Atmospheric Research “Formaldehyde Air Quality Assessment” Report C/0928 for NICNAS May 2004 be replicated. The CSIRO scenario used the annual national pollutant inventory data for a large wood products plant from year 2000. The formaldehyde emissions to air were assumed by CSIRO to be split, 50 percent each between stack and fugitive sources, for an assumed flat ground site location. The AWPA further requested a second modelling scenario as described in the following report. This model considers the same quantity of annual formaldehyde emissions being emitted from both stack and fugitive sources as the CSIRO model. However a more representative site source emission profile has been used in the model by distributing the various source emissions as follows:
Attachment 1 Air Dispersion Model Inputs compares the CSIRO Atmospheric Research “Formaldehyde Air Quality Assessment” Report C/0928 for National Industrial Chemicals Notification and Assessment Scheme (NICNAS) May 2004 inputs with the EML Air typical MDF plant scenario input data. Total annual emissions from stack and fugitive sources of 27,000 kg p.a. are identical in both cases. However the distribution is dissimilar because the CSIRO report splits the emission 50 percent each between the two (2) sources whereas the EML plant scenario for stack emissions is proportioned across the typical MDF plant (heat plant; dryers; press hoods & baghouses) based on actual stack tests. These stack sources represent 95.7 percent of the total emissions. For the EML volume sources the emission is the difference between total annual emission and stack source total emission; i.e. only 4.3 percent of the total emission. The CSIRO modelling assumed the stack and fugitive sources to be adjacent whereas the EML scenario assumed a spread of the sources as shown in a site layout (not to scale) Attachment 2. Stack heights; diameters; discharge temperatures and discharge velocities vary markedly as shown in Attachment 1. Again the EML scenario is based on test results for these sources. Similarly the EML scenario volume sources are assumed to be emitting from doors and roof ridge rather than a single source in the CSIRO model. The EML model assumes a building present whereas the CSIRO model has not considered the specific influence of wake effects on the dispersion from and around a typical large plant building. The receptor system is the same in both cases with a discrete receptor placed 100 metres from the sources in an easterly direction. The EML Air scenario places this receptor 100 metres from the eastern end of the building where fugitive and baghouse emissions are prominent and press hood emissions will be entrained in the building wake. Both models use the meteorological file used by CSIRO (CSIRO Atmospheric Research personal communication). To assist interpretation of the modelling predictions the EML Air scenario includes three (3) source groups i.e. all sources, stack sources only, and the fugitive sources only.
The predicted ground level concentrations (or PEC`s as referred to in the CSIRO report) are summarised in Table A17-1 below. Table A17-1: Predicted ground level concentration
Note * ppbv = (mg/m3 ÷ 1.2) x 1000 @ 300C same as CSIRO report. In particular the results in Table A17-1 can be summarised as follows:
iii. The EML Air typical plant scenario illustrates that stack and fugitive sources are predicted to be contributing almost equally to the predicted concentration.
The re-modelling of the CSIRO scenario for a large MDF plant emitting 27,000 kilograms per annum of formaldehyde has verified the predicted environmental concentrations reported to NICNAS for a large source. Using emission test results from the stack sources typical of a large MDF plant and:
- the modelling predicted environmental concentrations from a scenario plant configuration are an order of magnitude lower when compared to the CSIRO model (as shown in Table A17-1 above). Regardless of modelling using the typical test and plant configuration these predicted concentrations are 3 times higher than measurements at the boundary of two (2) wood panel facilities in Australia who reported < 10 ppb 24-hour average. Geoff White Senior Environmental Consultant References Victoria Government Environment Protection Act “The State Environment Protection Policy (Air Quality Management) (SEPP (AQM))” December 2001. ATTACHMENT 1 AUSTRALIAN WOOD PANELS ASSOCIATIONMDF PLANTS FORMALDEHYDE AIR DISPERSION AIR DISPERSION MODEL INPUTS
Notes: Plant scenario volume source emission based the following – Total formaldehyde emission 0.860 g/sec. Total stack source test emission 0.823 g/sec Difference as fugitive 0.037 g/sec 40% assigned to doors 0.015 g/sec 60% assigned to roof ridge 0.022 g/sec
Emissions Heat plant, Dryers, Presses and Baghouses based on MDF plant 2005 tests Building 100 x 25 x 12 Press hood vents 10 off 3 metre above roof height Formaldehyde 307 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:
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
