Title of Example

Example

To fulfil the EU Directives the UK Government has required all Local Authorities to undergo a staged review and assessment of air quality. As part of stage three, computer modelling of the main pollutants was undertaken. Bristol used the model ADMS –Urban produced by CERC for this process. This example is part of the process which modelled the point industrial sources.

Meteorology

Meteorological data were obtained from Trinity Consultants for 1998. These data were used for the 1998 validation model run and for the 2005 \ 2004 model run. This was the only meteorological data set used, due to time constraints and the lack of suitable PM₁₀ background data for the purposes of PM₁₀ source apportionment. Some sensitivity analysis was conducted at point receptors for validation purposes using 1997 meteorological data. Although Bristol City Council operates meteorological stations around the city, not all parameters required by ADMS – Urban are measured at these stations.

Atmospheric Chemistry for Nitrogen Dioxide

The calculation of nitrogen dioxide concentrations from emissions of NOx was done using the Generic Reaction Scheme feature in ADMS – Urban. This feature is fully explained in section 4.5.2 of the ADMS – Urban user manual (CERC, 1999), but in summary, comprises a set of eight chemical atmospheric reactions that use background monitored data to calculate NO ₂ concentrations. The background data used in this exercise is ratified data from AEE Harwell for 1998.

GIS Data “cleaning” Functions used on Model Output

The “grid” files generated by ArcView Spatial Analyst during interpolation of the ADMS – Urban “glt” (Grid Long Term) files are raster data sets derived from the IDW (Inverse Distance Weighted) algorithm. These data sets contain anomalous cells that hinder the visual interpretation of maps classified by concentration bands of air pollution. Data “cleaning” routines were therefore used to remove these anomalous cells and to “smooth” the boundaries. This is common practice in raster data processing in GIS, and although the output is changed somewhat, the integrity of the data is maintained.

Specifically these routines involved:

¨ Initial re-classification of the grid as an integer grid (necessary for the application of data cleaning routines)

¨ Application of the MajorityFilter request to remove outlying spurious data points

¨ Application of the BoundaryClean request to smooth the boundary of the exceedence area

¨ Re-application of the MajorityFilter request to remove residual outlying spurious data points

These functions remove “noise” from the model output and improve visual interpretation of the images. The errors introduced are minimal. Further details on the operation of these functions is available in ArcView help files and documentation.

The grid data sets are then converted to vector data (ArcView shapefiles) such that they can be overlaid on a raster map.

^{Sources Modelled at Stage Three}

Two maps showing the locations and types of sources modelled for the stage three assessment are shown below. Only NOx sources are shown here. The maps for PM₁₀ sources are very similar.

The sources for the purposes of modelling are broken down into three categories, grid, road (line) and point sources. Descriptions of how the emissions from each category are derived and calculated are contained in the following sections.

ãCrown copyright. All rights reserved: Bristol City Council LA09055L2000

· Figure 1 All NOx sources modelled for 2005

Grid Sources

Grid sources are used in ADMS – Urban to represent aggregated sources that need not be modelled explicitly as point or line sources. This helps to speed up run time and to enable the “intelligent gridding” option. Grid sources can also be used to represent “background” emissions, i.e. domestic heating, small industry etc.

ADMS – Urban automatically subtracts explicitly modelled sources from the grid sources, so even the explicitly modelled sources must be included in the grid sources.

For the purposes of this study, the grid sources used in the validation model run and the predictive model run were a combination of background source data and road and point sources that were not modelled explicitly.

Background Contribution

The background data were taken from the National Atmospheric Emissions Inventory (NAEI) data set available on the Internet. The area of each grid square was 1 square kilometre.

NAEI data is for 1996. There is an assumption that there will be negligible change in the nature and amount of these emissions. The data from the NAIE are broken down by sector, i.e. roads, small industry, domestic and other. All sectors apart from roads were summed and the emissions from these are used as the background sources.

The data files for NOx and PM₁₀ were processed in a spreadsheet such that the emissions were converted to grams per second per square kilometre from tonnes per year. The area of interest was extracted using GIS functions.

Point Source Contribution to Grid Sources

Because ADMS-Urban also subtracts emissions from point sources from the relevant grid square, it is necessary to add the point source emissions to the grid square prior to the model run.

Point Sources

The point sources used in the model for 2005 and 2004 are mainly operating part A processes under the control of the Environment Agency’s Integrated Pollution Control (IPC) regime. The master database for point sources included in the model is contained in an ADMS – Urban emissions inventory. The database includes sources and emissions from the Bristol area point sources and sources from the outside Bristol.

The following processes are not yet operating, but we have included them in the model run as we have been informed by the Agency or by the prospective operator that they will be running in the year of interest.

· Figure 2 Plant not currently operating but included in 2004 \ 2005 model

Enron Ltd has proposed a new 1200 MW gas fired power plant for the Severnside area. This plant will be approximately 300 metres from the existing 1200MW at Seabank. The Enron plant has not been included in the model, as we were notified after modelling work had started. The new plant will add to the concentrations already experienced on Severnside and may increase the frequency of exceedence of the hourly mean NAQS standard for nitrogen dioxide.

For all plant that is currently operating, Bristol City Council have liased closely with the relevant IPC Inspector to obtain the best estimates of emissions for 2005 and 2004. The operator in considering abatement plant required to be fitted by the Environment Agency over the next few years derives these estimates. The estimates of emissions in the target year include data not necessarily included on the public register. Estimates of NOx emissions are expressed in terms of tonnage of NOx as NO₂. Emissions of concentrations of NO₂ from industrial processes are likely to be over – predicted by modelling.

Improvements in the emissions from existing sources at Sevalco due to new abatement plant have been included in the modelling. Emissions from the new CHP plant itself, however, were not included in the modelling as the final specification for the plant was not complete at the time of modelling. The Environment Agency have required emissions of NOx from the plant to be no greater than those currently authorised. Modelling work is currently underway by contractors to Sevalco and this data will be incorporated in future review and assessment exercises.

We have modelled only plant that operates continuously. This is defined as over 8000 hours per year. It is not possible, at this stage, to model plant that operates in a transient and ad-hoc manner for future years. The point source modelling may therefore under predict in this respect. The most significant NOx source that operates in a transient manner is the British Gas Transco compressing plant near Hallen.

Part B processes were not modelled. The Part B processes emitting pollutants of concern operate in a transient manner and cannot be modelled in conjunction with continuously operating processes.

Although many of the sources are in Bristol, we have also included the plant outside the city boundary that may have an effect on air quality in the city due to long range transport. These sources are not included in grid sources.

Fugitive Sources of PM₁₀

In accordance with guidance in Chapter 8 of TG4 (00) we have examined information relating to fugitive sources of dust from industrial processes. The significant source in this respect is the zinc smelting plant, Britannia Zinc Ltd. in Avonmouth. A number of ore stockpiles, conveyor plants and vehicle movements around the site contribute to fugitive emissions from this plant. Control measures are in place to minimise fugitive dust. The difficulty in modelling fugitive dust is well known. Britannia Zinc operates a comprehensive monitoring program, and data from this have been used to assess the impact of fugitive dust on overall PM₁₀ in the locale.

The complete report for stage 3 and stage 4 of the Review and Assessment of air quality in Bristol can be found at www.bristol-city.gov.uk/airquality

Last Updated

13th January 2005