EcoSense

An Integrated Environmental Impact Assessment Model

1. THE ECOSENSE MODEL

1.1 Introduction

Taking into account this background, the objectives for the development of the EcoSense model were:

• to provide a tool supporting a standardised calculation of fuel cycle externalities,
• to integrate relevant models into a single system,
• to provide a comprehensive set of relevant input data for the whole of Europe,
• to enable the transparent presentation of intermediate and final results, and
• to support easy modification of assumptions for sensitivity analysis.

1.2 Scope of the EcoSense model

EcoSense was developed to support the assessment of priority impacts resulting from the exposure to airborne pollutants, namely impacts on health, crops, building materials, forests, and ecosystems. Although global warming is certainly among the priority impacts related to air pollution, this impact category is not covered by EcoSense because of the very different mechanism and global nature of impact. Priority impacts like occupational or public accidents are not included either because the quantification of impacts is based on the evaluation of statistics rather than on modelling. Version 2.0 of EcoSense covers 13 pollutants, including the 'classical' pollutants SO₂, NO_x, particulates and CO, as well as some of the most important heavy metals and hydrocarbons, but does not include impacts from radioactive nuclides.

1.3 The EcoSense Modules

Figure 1 shows the modular structure of the EcoSense model. All data - input data, intermediate and final results - are stored in a relational database system. The two air quality models integrated in EcoSense are stand-alone models, which are linked to the system by pre- and postprocessors. There are individual executable programs for each of the impact pathways, which make use of common libraries. The following sections give a more detailed description of the different EcoSense modules.

Reference Technology Database

Figure 1 Structure of the EcoSense model

Reference Environment Database

The reference environment database is the core element of the EcoSense database, providing data on the distribution of receptors, meteorology as well as a European wide emission inventory. All geographical information is organised using the EUROGRID co-ordinate system, which defines equal-area projection gridcells of 10 000 km² and 100 km² (Bonnefous a. Despres, 1989), covering all EU and European non-EU countries.

Data on population distribution and crop production are taken from the EUROSTAT REGIO database, which in some few cases have been updated using information from national statistics. The material inventories are quantified in terms of the exposed material area from estimates of 'building identikits' (representative buildings). Surveys of materials used in the buildings in some European cities were used to take into account the use of different types of building materials around Europe. Critical load maps for nitrogen deposition are available for nine classes of different ecosystems, ranging from Mediterranean scrub over alpine meadows to tundra areas. To simplify access to the receptor data, an interface presents all data according to administrative units (e.g. country, state) following the EUROSTAT NUTS classification scheme. The system automatically transfers data between the grid system and the respective administrative units.

In addition to the receptor data, the reference environment database provides elevation data for the whole of Europe on the 10x10 km grid, which is required to run the Gaussian plume model, as well as meteorological data (precipitation, wind speed and wind direction) and a European-wide emission inventory for SO₂, NO_x and NH₃ from EMEP 1990 which has been transferred to the EUROGRID-format.Exposure-Response Functions Using an interactive interface, the user can define any exposure-effect model as a mathematical expression. The user-defined function is stored as a string in the database, which is interpreted by the respective impact assessment module at runtime. All exposure-response functions compiled by the various 'area experts' of the ExternE Maintenance Project are stored in the database. Monetary Values The database provides monetary values for most of the impact categories following the recommendations of the ExternE economic valuation task group. In some cases there are alternative values to carry out sensitivity analysis

1.3.2 Air Quality Models

To cover different pollutants and different scales, EcoSense provides two air transport models completely integrated into the system:

• The Industrial Source Complex Model (ISC) is a Gaussian plume model developed by the US-EPA (Brode and Wang, 1992). The ISC is used for transport modelling of primary air primary air pollutants (SO₂, NO_x, particulates) on a local scale.
• The Windrose Trajectory Model (WTM) is a user-configurable trajectory model based on the windrose approach of the Harwell Trajectory Model developed at Harwell Laboratory, UK (Derwent, Dollard, Metcalfe, 1988). For current applications, the WTM is configured to resemble the atmospheric chemistry of the Harwell Trajectory Model. The WTM is used to estimate the concentration and deposition of acid species on a European wide scale.

1.3.3 Impact Assessment Modules

The impact assessment modules calculate the physical impacts and - as far as possible - the resulting damage costs by applying the exposure-response functions selected by the user to each individual gridcell, taking into account the information on receptor distribution and concentration levels of air pollutants from the reference environment database. The assessment modules support the detailed step-by-step analysis for a single endpoint as well as a more automised analysis including a range of prespecified impact categories.

1.3.4 Presentation of Results

Input data as well as intermediate results can be presented on several steps of the impact pathway analysis in either numerical or graphical format. Geographical information like population distribution or concentration of pollutants can be presented as maps. EcoSense generates a formatted report with a detailed documentation of the final results that can be imported into a spreadsheet programme.

1.4 The air quality models integrated in EcoSense

1.4.1 Local scale modelling of primary pollutants - the Industrial Source Complex model Close to the plant, i.e. at distances of some 10-50 km from the plant, chemical reactions in the atmosphere have little influence on the concentrations of primary pollutants, if NO and its oxidised counterpart NO₂ can be summarised as NO_x. Due to the large emission height on top of a tall stack, the near surface ambient concentrations of the pollutants at short distances from the stack are heavily dependent on the vertical mixing of the lower atmosphere. Vertical mixing depends on the atmospheric stability and the existence and height of inversion layers (whether below or above the plume). For these reasons, the most economic way of assessing ambient air concentrations of primary pollutants on a local scale is a model which neglects chemical reactions but is detailed enough in the description of turbulent diffusion and vertical mixing.

An often used model which meets these requirements is the Gaussian plume model. The concentration distribution from a continuous release into the atmosphere is assumed to have a Gaussian shape: 
 

where:	c(x,y,z)	concentration of pollutant at receptor location (x,y,z)
	Q	pollutant emission rate (mass per unit time)
	u	mean wind speed at release height
		standard deviation of lateral concentration distribution at downwind distance x
		standard deviation of vertical concentration distribution at downwind distance x
	h	plume height above terrain

The assumptions embodied into this type of model include those of idealised terrain and meteorological conditions so that the plume travels with the wind in a straight line. Dynamic features which affect the dispersion, for example vertical wind shear, are ignored. These assumptions generally restrict the range of validity of the application of these models to the region within some 50 km of the source. The straight line assumption is rather justified for a statistical evaluation of a long period, where mutual changes in wind direction cancel out each other, than for an evaluation of short episodes.

EcoSense employs the Industrial Source Complex Short Term model, version 2 (ISCST2) of the U.S. EPA (Brode and Wang, 1992). The model calculates hourly concentration values of SO₂, NO_x and particulate matter for one year at the center of each small EUROGRID cell in a 10 x 10 grid centred on the site of the plant. Effects of chemical transformation and deposition are neglected. Annual mean values are obtained by temporal averaging of the hourly model results.

The s_y and s_z diffusion parameters are taken from BMJ (1983). This parameterisation is based on the results of tracer experiments at emission heights of up to 195 m (Nester and Thomas, 1979). More recent mesoscale dispersion experiments confirm the extrapolation of these parameters to distances of more than 10 km (Thomas and Vogt, 1990).

The ISCST2 model assumes reflection of the plume at the mixing height, i.e. the top of the atmospheric boundary layer. It also provides a simple procedure to account for terrain elevations above the elevation of the stack base:

• The plume axis is assumed to remain at effective plume stabilisation height above mean sea level as it passes over elevated of depressed terrain.
• The effective plume stabilisation height h_stab at receptor location (x,y) is given by:

 

where:	h	plume height, assuming flat terrain
	hs	height of the stack
	zs	height above mean sea level of the base of the stack
		height above mean sea level of terrain at the receptor location

• The mixing height is terrain following.

It is the responsibility of the user to provide the meteorological input data. These include wind direction, wind speed, stability class as well as mixing height, wind profile exponent, ambient air temperature and vertical temperature gradient.

1.4.2 Regional scale modelling of primary pollutants and acid deposition - the Windrose Trajectory Model

With increasing distance from the stack the plume spreads vertically and horizontally due to atmospheric turbulence. Outside the area of the local analysis (i.e. at distances beyond 50 km from the stack), it can be assumed for most purposes that the pollutants have vertically been mixed throughout the height of the mixing layer of the atmosphere. On the other hand, chemical transformations can no longer be neglected on a regional scale. The most economic way to assess annual, regional scale pollution is a model with a simple representation of transport and a detailed enough representation of chemical reactions.

The Windrose Trajectory Model (WTM) used in EcoSense to estimate the concentration and deposition of acid species on a regional scale was originally developed at Harwell Laboratory by Derwent and Nodop (1986) for atmospheric nitrogen species, and extended to include sulphur species by Derwent, Dollard and Metcalfe (1988). The model is a receptor-orientated Lagrangian plume model employing an air parcel with a constant mixing height of 800 m moving with a representative wind speed. The results are obtained at each receptor point by considering the arrival of 24 trajectories weighted by the frequency of the wind in each 15° sector. The trajectory paths are assumed to be along straight lines and are started at 96 hours from the receptor point. The chemical scheme of the model is shown in Figure 2.

In EcoSense, the model is implemented by means of

• a set of parameters and chemical equations in the Ecosense database which defines the model
• a model interpreter (wmi.exe)
• a set of meteorological input data (gridded wind roses and precipitation fields) in the reference environment database
• emission inventories for NO_x, SO₂ and ammonia, which are also provided in the reference environment database
• additional emissions of the plant from the reference technology database

References

Bonnefous, S. and A. Despres (1989): Evolution of the European data base, IPSN/EURATOM - CEA Association, BP 6, 92265 Fontenay-Aux-Roses, France.

BMJ (1983): Der Bundesminister der Justiz (ed.). Störfall-Leitlinien. Bundesanzeiger 35, 245a.

R.W. Brode, J. Wang: Users's Guide for the Industrial Source Complex (ISC2) Dispersion Models Volumes I-III. EPA-450/4-92-008a. EPA-450/4-92-008b. EPA-450/4-92-008c. U.S. Environmental Protection Agency, 1992, Research Triangle Park, North Carolina 27711.

Derwent, R.G. and K. Nodop (1986): Long-range transport and deposition of acidic nitrogen species in north-west Europe. Nature 324, 356-358.

R.G. Derwent, G.J. Dollard, S.E. Metcalfe: On the nitrogen budget for the United Kingdom and north-west Europe. Q. J. R. Meteorol. Soc. 114, 1127-1152, 1988.

Hollingsworth, A. (1987): Objective analysis for numerical weather prediction. Collection of papers presented at The WMO/IUGG NWP Sympsium, Tokyo, 4-8 August 1986, 11­59.

Nester, K. and P. Thomas (1979): Im Kernforschungszentrum Karlsruhe experimentell ermittelte Ausbreitungsparameter für Emissionshöhen bis 195 m. Staub 39, 291-295.

Nordeng, T.E. (1986): Parameterization of physical processes in a three-dimensional numerical weather prediction model. Technical Report No. 65. DNMI, Oslo.

Sandnes, H. and H. Styve (1992): Calculated Budgets for Airborne Acidifying Components in Europe 1985, 1987, 1988, 1989, 1990 and 1991. EMEP/MSC-W Report 1/92, Oslo.

Thomas, P. and S. Vogt (1990): Mesoscale Atmospheric Dispersion Experiments Using Tracer and Tetroons. Atmospheric Environment 24a, 1271-1284.