Guidelines on Assessing the 'TOTAL' Air Quality Impacts

1. Total Impacts - 3 Major Contributions

1.1 In evaluating the air quality impacts of a proposed project upon air sensitive receivers, contributions from three tiers of emission sources, depending on their distance from the site, should be considered. These are:

Primary contributions:	project induced
Secondary contributions:	pollutant-emitting activities in the immediate neighbourhood
Background contributions:	pollution not accounted for by the previous two tiers

2. Nature of Emissions

In most cases, the project-induced emissions are fairly well defined and quite often (but not necessarily) the major contributor to local air quality impacts. Examples include those due to traffic network, building or road construction projects.

Within the immediate neighbourhood of the project site, there are usually pollutant emitting activities contributing further to local air quality impacts. For most local scale projects, any emission sources in an area within 500m of the project site boundary with notable impacts should be identified and included in an air quality assessment to cover the short-range contributions. In the exceptional cases where there is one or more significant sources nearby, the study area may have to be extended or alternative estimation approach employed to ensure these impacts are reasonably accounted for.

The above two types of emission contributions should account for, to a great extent, the air quality impacts upon local air sensitive receivers, which are often amenable to estimation by the ‘steady state dispersion' type of models. However, a background air quality level should be prescribed to indicate the baseline air quality in the region of the project site, which would account for any pollution not covered by the two preceding contributions. The emission sources contributing to the background air quality would be located further afield and not easy to identify. In addition, the transport mechanism by which pollutants are carried over long distances is rather complex and cannot be adequately estimated by the 'steady state dispersion' type of models.

Background contributions are usually accounted for by the spatially averaged concentrations of regional air quality models (e.g. PATH) at a certain spatial resolution (e.g. 1km x 1km). The impact due to a point source, on the other hand, is usually confined within a plume whose dimension varies with downwind distance. When a point source is located close to a receiver such that its spatially varying contribution cannot be well represented by the spatially averaged contribution of a regional air quality model, it should be modelled by a dispersion model; however, when obstacles are present so that their mixing effect can even out the point source's spatially varying contribution, the point source can then be accounted for by a regional air quality model. In general, a major point source located within 4km from a receiver should be evaluated to see if it can cause direct impact and to establish its proper modelling tool accordingly (i.e. using dispersion model or regional model). The major point sources need to be considered in an air quality assessment and the associated details can be found [Here] for easy reference.

3. Estimation Approach

3.1.1 Balancing the need for representativeness with the resource demand for long-term simulations, EPD is currently accepting a minimum of one-year simulation of air quality for environmental impact assessment.

3.1.2 The default approach (see Additional Note at the end for details) entails both meteorological and air quality data for all relevant time scales to be extracted from the PATH(or any approved) modelling system on an hour-by-hour basis. (A brief description of PATH is in section 3.2 and more detailed description can be found on http://www.epd.gov.hk/epd/english/environmentinhk/air/guide_ref/guide_aqa_model.html.) Meteorological data driving the air quality models for assessing Tiers 1 and 2 emission sources and the concentration data to serve as Tier 3 contribution should be extracted consistently from the same grid cell(s) which cover(s) the project's Tier 1 and Tier 2 assessment areas.

3.1.3 Over the spatial scale of typical EIAs in Hong Kong, the transport of pollutants from source to receiver for Tier 1 and Tier 2 usually happens in an hour or less. Such dispersion situation is also represented by the models for Tiers 1 and 2 which use hourly inputs to produce hourly outputs. Longer time averaged concentrations (e.g. daily, annual … etc.) should thus be constructed using these hourly values as basic building blocks.

3.1.4 The meteorological and air quality data pertaining to each of the three tiers must be fully synchronised to produce credible results, e.g. the daily ‘Total’ air quality would be the sum of the contributions from the three tiers for the same day.

3.1.5 Approaches other than the default approach to derive meteorological and air quality data (including those based on observation data only) will have to be fully justified to EPD on a case-by case basis (see Section 3.4.1). In any case, use of short term measurements to prescribe the background concentrations is not acceptable.

3.2.1 Using air pollutant concentration outputs from the PATH modelling system as the Tier 3 background contribution has advantage over other sources of information since the system is capable of accounting for the changes to future air quality resulting from technology changes and deliberate government policies to improve air quality, both in Hong Kong and the surrounding areas, if the emission input to PATH is properly represented.

3.2.2 The PATH system consists of three major modules: a meteorological driver, an emission module and a chemical transport model. These are supported by the corresponding input data (see schematics in Figure 1).

Figure 1: Schematics of the PATH system

3.2.3 PATH is set up on a three-dimensional grid system with horizontal nesting. The most commonly used horizontal domain currently has grid spacing of 1 km in both the N-S and E-W direction (see Figure 2). The thickness of the first model layer can be taken to be 17 m. The relevant outputs of the PATH systems are: gridded meteorological and pollutant concentration data every hour of the simulation period.

Figure 2: PATH's horizontal domain with grid spacing of 1 km

3.2.4 EPD has made available model outputs for public access (http://www.epd.gov.hk/epd/english/environmentinhk/air/guide_ref/guide_aqa_model.html ). These include:

• Meteorological data for one representative year appropriate to drive air quality simulations using PATH, and

• Hourly concentrations over the first ten layers of the PATH domain for NO, NO₂, SO₂, RSP, FSP, O₃ and CO for some future years. These concentrations are linearly interpolated between model outputs for at least 2 years which are generated using EPD's best estimate / projection of emissions and the meteorological data for the representative year.

3.2.5 Air quality modelling systems other than PATH may be used if EPD can be convinced that the proposed system is needed for the situation in question and / or is superior to PATH (see Guidelines on the Use of Alternative Computer Models in Air Quality Assessment).

3.3.1 Since PATH's concentration outputs from EPD have accounted for Tier 2 emissions, though in a different way to Tier 2 simulations because PATH is a grid-based model whereas air quality models for assessing Tier 2 sources are commonly of the plume / puff type, there will be double-counting of emissions in the total simulation if such PATH's concentrations are used unadjusted. If this double-counting is undesirable for the assessment, then a re-simulation of the Tier 3 contribution can be made with the emissions already accounted for in Tier 2 removed from the corresponding grid(s) and the re-simulated concentrations used as Tier 3 contribution. A similar logic would apply to using other chemical transport models (if approved by EPD) for Tier 3 contributions.

3.3.2 Care must be taken to ensure that all concentrations (NO₂, O₃, RSP …etc.) from Tier 3 are taken consistently from either the original run (with Tier 2 emissions included) or the re-generated run (with Tier 2 emissions excluded to avoid double-counting). Any deviation from this basic consistency should be made clear, fully justified to and approved by EPD and documented in detail in appropriate reports.

3.4.1 The table below summarizes the possible combinations of input data for conducting a ‘Total’ air quality assessment. They are arranged in ascending order (from top to bottom) of computer resource requirement to facilitate finding an appropriate level of assessment.

3.4.2 In this table, three types of data are referred to. Within this context, meteorological data are always for driving air quality model (i.e. dispersion models) for assessing Tiers 1 and 2 emission sources; emissions always refer to those used in a photochemical model like PATH; and concentration data always refer to Tier 3 background contribution.

These should be established by the project proponent to the satisfaction of EPD

** See Section 3.2.4 on the meteorological and concentration data available for public access from EPD.

 

Modelling Section, Air Science and Modelling Group
Environmental Protection Department
Revised: January 2023

Additional Note Model-output-based and Observation-based Approach to Air Quality Assessment

In the past, before EPD made meteorological and air quality model outputs publicly available, two approaches in driving models for air quality assessment were used and equally accepted: one based solely on observations and the other based solely on model outputs. The only requirement was that the two approaches would have to be carried out self-consistently, meaning that using both model outputs and measurements to drive the models for assessment is not allowed within one assessment.

Using model outputs only for air quality assessment bypasses the need to justify that future background air quality would be an improvement over the present situation that is needed for the observation-based approach (see section 3.2.1). Due to the convenience of using model outputs for this 3-tier assessment, almost all air quality assessments in EIAs have now been made using model outputs. Given this trend, EPD is now considering model-output-based assessment as the default approach.