Nimba Western Area Iron Ore Concentrator Mining Project, Liberia. Environmental and Social Impact Assessment

Transcription

1 Nimba Western Area Iron Ore Concentrator Mining Project, Liberia Environmental and Social Impact Assessment Volume 3, Part 5: Air Quality Impact Assessment March 2013 Prepared for: ArcelorMittal Liberia Limited

2 REVISION SCHEDULE Rev Date Details Prepared by Reviewed by Approved by 1 28 th May 2012 First Draft Report Danny Duce Principal Air Quality Consultant Garry Gray Associate Gareth Hearn Technical Director 2 20 th July 2012 Second Draft Report Including power plant dispersion model Danny Duce Principal Air Quality Consultant Garry Gray Associate 3 13 th September October 2012 Third Draft Report Updates to power plant dispersion model design Draft Final Danny Duce Principal Air Quality Consultant Danny Duce Principal Air Quality Consultant Chuansen Ren Senior Air Quality Specialist Tanya Romanenko Gareth Hearn Tanya Romanenko Project Manager Martin Edge Director Europe & Africa 5 March 2013 Final Danny Duce Principal Air Quality Consultant Tanya Romanenko Project Manager Martin Edge Director Europe & Africa URS Scott House Alençon Link Basingstoke Hampshire RG21 7PP Tel +44 (0) Fax +44 (0)

3 Limitations URS has prepared this Report for the sole use of ArcelorMittal Liberia Limited ("Client") in accordance with the Agreement under which our services were performed and may not be relied upon by any other party without the prior and express written agreement of URS. No other warranty, expressed or implied is made as to the professional advice included in this Report or any other services provided by URS. The methodology adopted and the sources of information used by URS in providing its services are outlined in this Report. The work described in this Report was undertaken variously between 2010 and 2011 for Phase 1 and 2011 to June 2012 for Phase 2 and is based on the conditions encountered and the information made available by the Client during the said period of time. The scope of this Report and the services are accordingly factually limited by these circumstances. The layouts shown in supporting Volumes 3 to 5 inclusive have been superseded. The final layouts (December 2012) are provided and assessed in Volume 1 only. Where assessments of works or costs identified in this Report are made, such assessments are based upon the information available at the time and where appropriate are subject to further investigations or information which may become available. URS disclaim any undertaking or obligation to advise any person of any change in any matter affecting the Report, which may come or be brought to URS attention after the date of the Report. Certain statements made in the Report that are not historical facts may constitute estimates, projections or other forwardlooking statements and even though they are based on reasonable assumptions as of the date of the Report, such forward-looking statements by their nature involve risks and uncertainties that could cause actual results to differ materially from the results predicted. URS specifically does not guarantee or warrant any estimate or projections contained in this Report. Where field investigations are carried out, these have been restricted to a level of detail required to meet the stated objectives of the services. The results of any measurements taken may vary spatially or with time and further confirmatory measurements should be made after any significant delay in issuing this Report. Copyright This Report is the copyright of the Client. Any unauthorised reproduction or usage by any person other than the Client is strictly prohibited. 2

4 ENVIRONMENTAL AND SOCIAL IMPACT ASSESSMENT REPORT STRUCTURE Section Report Title Volume 1 Main (ESIA) Report Volume 2 Legal and Administrative Framework Volume 3 Physical Environment: Baseline Conditions (Supplement to Phase 1 Baseline Only) and Impact Assessments Volume 3, Part 1.1 Geo-Mapping Volume 3, Part 1.2 Geology, Soils and Land Use Volume 3, Part 1.3 Assessment of No Way Camp Disaster and General Considerations for Nimba Western Area Iron Ore Concentrator Mining Project Volume 3, Part 1.4 Review of Slope Stability and Drainage Conditions During Phase 1 Operations at the Mine Sites Volume 3, Part 1.5 Review of Slope Stability and Erosion Along the Railway Volume 3, Part 2 Groundwater Baseline and Impact Assessment Volume 3, Part 3.1 Hydrology Baseline and Impact Assessment Volume 3, Part 3.2 Geochemistry and Water Quality (Focusing on The Potential for Acid Rock Drainage (ARD) Formation Arising From Ore, Tailings and Waste Rock Materials) Volume 3, Part 4 Climate Change Scenarios by Met Office (UK) Volume 3, Part 5 Air Quality Impact Assessment Volume 3, Part 6 Noise and Vibration Impact Assessment Volume 3, Part 7 Landscape Character and Visual Amenity Impact Assessment Volume 4 Biological Environment: Baseline Conditions (Supplement to Phase 1 Baseline Only) and Ecological Impact Assessment Volume 4, Part 1.1 Forest Botanical Impact Assessment Volume 4, Part 1.2 Grassland Botanical Impact Assessment Volume 4, Part 2 Zoological Impact Assessment, Terrestrial and Coastal and Marine Volume 4, Part 3 An assessment of freshwater fish and crustacean consumption in Northern Nimba, Liberia Volume 4, Part 4 Bushmeat and Biomonitoring Studies in the Northern Nimba Conservation Area by Conservation International Volume 5 Socio-economic Environment: Baseline Conditions (Supplement to Phase 1 Baseline Only) and Social Impact Assessment Volume 5, Part 1 Socio-Economic Baseline for Buchanan, Greenhill Quarry and Areas in Nimba that will be Affected by TMF Operations. Volume 5, Part 2 Social Impact Assessment, and Framework Social Management Plan Volume 5, Part 3 Cultural Heritage Assessment ARCELORMITTAL LIBERIA LTD ENVIRONMENTAL AND SOCIAL MANAGEMENT PLANNING DOCUMENTATION STRUCTURE Volume 6 Environmental Management Planning Volume 6, Part 1 Framework Resettlement Action Plan for Phase 2 Volume 6, Part 2 ArcelorMittal Liberia Environmental Standards Manual Volume 6, Part 3.1 Overall Environmental Management Plan for Phase 2 Volume 6, Part 3.2 Environmental Management Plan: Construction Works near Mount Tokadeh Volume 6, Part 3.3 Environmental Management Plan: Operation of Quarries Volume 6, Part 3.4 Environmental Management Plan: Rehabilitation of Facilities at the Port of Buchanan Volume 6, Part 3.5 Environmental Management Plan: Operation of the Buchanan-Tokadeh Railway Volume 6, Part 3.6 Environmental Management Plan: Operation of the Port of Buchanan, including Offshore Transhipment Volume 6, Part 3.7 Hazardous Materials and Waste Management Plan for Phase 2 Volume 6, Part 3.8 Townships Management Plan Volume 7 Framework of the Proposed Mine and Infrastructure Closure Plan for Phase 2 Volume 8 Framework of the Proposed Environmental Offset Programme for Phase 2 SUPPLEMENTARY INFORMATION TO PHASE 2 ENVIRONMENTAL AND SOCIAL IMPACT ASSESSMENT REPORT Volume 9 Assessment of Legacy Environmental Issues in the Former LAMCO Mines and Industrial Areas 3

5 TABLE OF CONTENTS 1 INTRODUCTION Overview Assessment Scope Overview Emissions of Dust and Fugitive Particulate Matter Emissions from the Mine Site Power Plant Stacks Greenhouse Gas Emissions Activities Omitted from the scope of the ESIA Spatial Scope Temporal Scope LEGAL AND POLICY CONTEXT Criteria and Standards to be Adopted for the ESIA METHODOLOGY Baseline Determination Air Quality Sensitive Receptors Baseline Dust and Particulate Matter Survey of Baseline Nitrogen Dioxide and Sulphur Dioxide Meteorological Conditions and Topography Prediction and Evaluation of Air Quality Effects Emissions of Dust and Particulate Matter Dispersion Modelling of Power Plant Emissions BASELINE CONDITIONS Receptor Locations Mine Area Railway Corridor Greenhill Quarry Port (Landside) Dust and Particulate Matter Sources of Baseline Dust and Particulate Matter Measured Dust and Particulate Matter Concentrations Measured Baseline Dust Deposition Rates Measured Baseline Nitrogen Dioxide and Sulphur Dioxide Greenhouse Gas Levels Meteorology Topography and Vegetation Ore and Surface Soil Characteristics OPERATIONAL PHASE IMPACTS

6 5.1 Fugitive Emissions of Dust and Particulate Matter from the Mine Area Sources of Dust Prediction of Effects on Human Populations Prediction of Effects on Crops and Sensitive Flora Emissions of Combustion Pollutants from Power Plant Stacks Railway Corridor Dust Port (Landside) Dust Effects Emissions of Greenhouse Gases MITIGATION Principles Specific Measures Mine Area Ore Processing and Concentrator Area Railway Corridor Port (landside) RESIDUAL EFFECTS Mine and Processing Areas Emissions of Combustion Pollutants from Power Plant Stacks Railway Corridor Greenhill Quarry Port (landside) Greenhouse Gas Emissions CONCLUSIONS REFERENCES APPENDIX 1: FIGURES APPENDIX 2: BUCHANAN POWER PLANT DISPERSION MODELLING ASSESSMENT

7 1 INTRODUCTION 1.1 Overview This report considers the potential effects of the Scheme arising from changes in air quality associated with dust generation and emissions of gaseous pollutants and fine particulate matter. Changes in air quality can impact on both human receptors and the wider environment, notably on plants of biodiversity value and crop productivity. Consideration is also given to the Scheme s contribution to greenhouse gas emissions. An air quality assessment of Phase 1 of the project, involving the production of up to 20 million tonnes of Direct Shipping Ore (DSO), rehabilitation and operation of the railway line to Buchanan and landside port operations was submitted as part of the Phase 1 ESIA in September This report extends the scope of that study to incorporate the proposed Phase 2 works, which include: production of up to 180 million tonnes of lower grade, beneficiated ore at Mounts Tokadeh, Gangra and Yuelliton between 2015 and 2026; construction and operation of an ore concentration plant, water abstraction system, dam and storage reservoir, tailings management facility and power plant; and further enhancements to rail, port and community infrastructure. 1.2 Assessment Scope Overview Receptors and resources considered in this study comprise: People in local settlements or in sensitive areas (hospitals and schools etc) or other areas (e.g. agricultural fields) subject to relevant exposure to dust or gaseous pollutants; Agricultural crops; Flora considered to be important biodiversity 1 ; and National greenhouse gas levels. Potential effects on such resources and receptors resulting from the Scheme comprise: effects on human health associated with increased levels of fine dust (particulate matter less than 10µm diameter); health effects due to inhalation of toxic dusts; effects on amenity associated with increased rates of dust deposition; effects on human health from gaseous pollutants; effects on plants and crops arising from a blocking of stomata 2 and/or reduction in photosynthetic activity as a result of dust deposition; and effects on greenhouse gas emissions at a national level. The potential air quality impacts of the Scheme are set out as within Table 1.1, below. 1 A definition of important biodiversity is provided in the Freshwater and Terrestrial Biodiversity report 2 Stomata: the pore openings on plant leaves that can open and close according to the metabolic needs of the plant. Ports for the exchange of oxygen and CO 2 for photosynthesis 6

8 TABLE 1.1: SCOPE OF AIR QUALITY ASSESSMENT Area Emission Activities Mining Area Railway Corridor Port Townships Fugitive emissions of dust and particulate matter from mining activity Emissions of gaseous pollutants and particulate matter from mobile plant, excavators and dump trucks Fugitive emissions of dust and particulate matter Emissions of gaseous pollutants and particulate matter Emissions of greenhouse gases Fugitive emissions of dust and particulate matter Emissions of gaseous pollutants and particulate matter Emissions of greenhouse gases Fugitive emissions of dust and particulate matter Emissions of gaseous pollutants and particulate matter Emissions of greenhouse gases Site preparation and construction works (including concentration plant, power plant, tailings management facility and water retention dam) Soil stripping and overburden removal Drilling and blasting Excavation of ore material Transportation of ore from mine to primary crusher dump hoppers Operation of crushing stations and transfer conveyor Operation of the crushed ore blending system Operations within crushed ore blending stockpile area Operation of concentration plant, including SAG mills, grinding screens, discharge to concentrate surge stockpile or load-out silo system Operation of rail load out facility and train loading system Excavation of ore material Transportation of ore from mine to primary crusher dump hoppers Emissions from open rail wagons during transport Emissions from vehicles using the service road Emissions from rail locomotives Emissions from vehicles on the service road Emissions from rail locomotives and road traffic on the service road Ore handling operations Emissions from ships Emissions from ships Use of unsurfaced roads Emissions from road vehicles Emissions from domestic fuel consumption and road vehicles Scoped In Scoped Out 7

9 1.2.2 Emissions of Dust and Fugitive Particulate Matter Dust emissions, dispersion patterns and impacts are difficult to predict due to the lack of reliable emissions factors for the wide range of Scheme activities that might give rise to dust. Emission factors can be derived from a number of sources including documents published by the United States Environmental Protection Agency (USEPA, 1998). However, there is significant uncertainty regarding many of these factors, particularly when they are applied to mining operations that differ to those from which the source data were collected. For these reasons, the approach adopted to predict and evaluate the effects of dust emissions within the ESIA has been based on a qualitative methodology, using information on: the likely level of dust generation based upon the type, location, duration and frequency of activities; size of the dust particles, and hence the likely distance they could travel; meteorological observation data on rainfall, wind direction, and topographical conditions (vegetation and topography) to determine likely dilution and dispersion characteristics; and distances to sensitive receptors. The assessment has used evidence from published reports and other relevant guidance, in conjunction with project specific observations undertaken during a site visit with regard to the distances over which significant impacts might occur Emissions from the Mine Site Power Plant Stacks The main air quality impact during the operation of the proposed power plant would be emissions to air from the combustion of fuel within the power plant engines. The primary fuel for the plant will be Heavy Fuel Oil (HFO). There are currently no national limits for emissions from power plants in Liberia. Therefore emission guidelines for new thermal power plants burning fossil fuels, as detailed in the IFC guidelines for Thermal Power Plants, have been employed in the design of the facility. The air quality impacts of the proposed power plant have been evaluated using the plume dispersion model ADMS, which is able to calculate maximum ground level concentrations at sensitive receptors close to the Scheme boundary. The dispersion model has been used to verify the proposed stack height as appropriate, and to demonstrate that the predicted impacts of the operation of the plant at air quality sensitive receptors are acceptable in the context of WHO air quality standards Greenhouse Gas Emissions Consideration has been given to the Scheme s contribution to national greenhouse gas emissions, on a qualitative basis. The working methods and procedures employed have been evaluated to consider the efficiency of proposed operations, taking into account the International Finance Corporation Performance Standard 3 (IFC, 2006) Activities Omitted from the scope of the ESIA The following activities have been omitted from the scope of the ESIA for the reasons provided below: Mine Areas Changes in air quality resulting from emissions of gaseous pollutants and fine particulate matter from on-site plant including mobile equipment and trucks. The numbers of such plant are low and it is assumed that they will comply with strict emissions standards. They have therefore been scoped out of ESIA on this basis. 8

10 Chemical analyses of samples of local soil were presented in the Phase 1 ESIA, and it was concluded that any potential toxic effects of dust emissions upon human health and ecosystems were considered unlikely. It is not therefore proposed to undertake any further detailed assessment of toxicity effects within the Phase 2 ESIA. Railway Corridor Emissions of gaseous pollutants and fine particulate matter from locomotives have been scoped out, as there would be a low frequency of movements, and the locomotives will comply with US Tier 2 emissions limits. Port (Landside) Emissions of gaseous pollutants and fine particulate matter from power generation and on-site plant for similar reasons as identified for the mine areas above. Townships Infrastructure Changes in air quality resulting from emissions of gaseous pollutants and fine particulate matter from road vehicle exhausts. The number of road vehicle movements is expected to be low, in comparison with those typical of urban areas. They have therefore been scoped out of ESIA on this basis Spatial Scope Gaseous pollutants, fine particulate matter and greenhouse gas emissions from airport operations as airports are scoped out of this ESIA and the anticipated number of aircraft movements is anyway insufficient to significantly affect air quality. Dust concentrations decline rapidly with increasing distance from the emissions source, due to the effects of deposition (for large dust particles) and atmospheric dilution and dispersion (for fine particles). The spatial extent for effects associated with dust has therefore been determined by such factors. Based on such considerations (as described in detail within section 4.2), the spatial scope is up to: One kilometre from the edge of the mine areas; 500 metres from the mine access and haul roads, stockpiles and loading areas; 200 metres from the railway line and service roads, and 200 metres from the port stockpiles. Impacts from the Tokadeh power plant have been considered at the closest sensitive receptors in each direction from the mine site. The receptors included in the study are all within 6 km of the power plant stacks. The IFC standard for greenhouse gases looks at the absolute contribution on a national scale, and so spatial scope is not relevant to this issue Temporal Scope Any effects on air quality and greenhouse gases will be confined to the operational lifetime of the scheme. 9

11 2 LEGAL AND POLICY CONTEXT Air quality standards, legislation, and guidance, which are considered to be of direct relevance to the assessment of the air quality impacts of the Phase 2 development are set out within this section; however this is not an exhaustive list. Part IV, Section 36, of the Environmental Protection and Management Law (EPML) of the Republic of Liberia (GoL, 2003) states that the Environmental Protection Agency (EPA) shall, in consultation with the relevant Line Ministry, establish criteria and procedures for the measurement of air quality, including ambient air quality standards. The Liberian Environmental Protection Agency subsequently published draft air quality regulations in September These criteria, however, are not yet in place and at the current time there are no other formally adopted national criteria that can be used to determine the significance of air quality impacts. In the absence of such formally adopted national standards, the Phase 2 ESIA will be undertaken as set out within the scoping report, in accordance with international best practice, notably World Health Organisation (WHO) air quality guidelines (WHO, 2000, 2005) for gaseous pollutants and fine particulate matter. These guidelines have been adopted for projects funded by the International Finance Corporation (IFC, 2007a), and also form the basis for standards adopted in the European Union. In the absence of international standards for dust deposition rates the action level used for residential areas in South Africa have been adopted. A consideration of the draft Liberian air quality regulations will also be made. Liberia is a signatory to both the United Nations Framework Convention on Climate Change (UNFCCC) (UN, 1992a) and the Vienna Convention (for the Protection of the Ozone Layer) (UNEP,1985), but has not signed the Kyoto Protocol (UN, 1992b); Liberia is, therefore, not legally bound to any international targets. However, as a matter of good practice this ESIA would include a consideration of the implications for direct contribution of the Scheme to greenhouse gas emissions, taking account of the International Finance Corporation Performance Standard 3 (IFC, 2006). Consideration would also be given to the following sector-specific guidance on air quality, relevant to proposed Phase 2 operations: Environmental, Health and Safety Guidelines for Mining published by the International Finance Corporation (IFC, 2007b); Environmental, Health and Safety Guidelines for Thermal Power Plants published by the International Finance Corporation (IFC, 2007c); and European Commission Best Available Techniques Reference Document on the production of Iron and Steel (EC, 2001). 2.1 Criteria and Standards to be Adopted for the ESIA In proposing air quality standards for the project, it is important to draw a distinction between the guidelines recommended by WHO, and the standards that have been brought into legislation by e.g. the European Union. The former are purely based on the scientific and medical evidence of the effects of an individual pollutant. The latter take into account the extent to which the standards are expected to be achieved by a certain date including economic efficiency, practicability, existing air quality conditions and technical feasibility and timescale. Based on international best practice, the air quality standards adopted for the ESIA are set out in Table 2.1 below. 10

12 TABLE 2.1: AIR QUALITY STANDARDS ADOPTED FOR THE ESIA Pollutant Dust deposition PM 10 PM 2.5 Nitrogen dioxide (NO 2 ) Sulphur dioxide Carbon monoxide Averaging Period Standard 30 days 600 mg/m 2 /day (not to be exceeded more than three times per year, no two sequential months) 24 hours 150 µg/m 3 (99 th percentile) Annual 70 µg/m 3 mean 24 hours 75 µg/m 3 (99 th percentile) Annual 35 µg/m 3 mean Standard Derived From South African action level for residential areas (SANS 1929, 2004) IFC (adopted from WHO Guidelines, Interim Target 1) 1 hour 200 µg/m 3 IFC (adopted from Annual 40 µg/m 3 WHO Guidelines) mean 10 min mean 500 µg/m 3 IFC (adopted from WHO Guidelines) 24 hours 125 µg/m 3 IFC (adopted from WHO Guidelines, Interim Target 1) 8 hours 10 mg/m 3 IFC (adopted from WHO Guidelines) Sources Mining operations, roads, agriculture and various nonanthropogenic sources Mining operations, vehicle exhausts, railway locomotives, power generation Vehicle exhausts, railway locomotives, power generation Railway locomotives, heavy fuel oil-fired power generation Vehicle exhausts, railway locomotives, power generation PM 10 and PM 2.5 is fine particulate matter with an aerodynamic diameter of less than 10 and 2.5 micrometres respectively 11

13 3 METHODOLOGY 3.1 Baseline Determination Air Quality Sensitive Receptors Human Receptors An extensive survey of settlements and villages in the area around the mine sites, railway corridor and port was undertaken in advance of Phase 1 of the project, during The identification of hamlets and individual properties potentially affected by emissions to air consisted of: Interpretation of aerial photography to identify dwellings within varying distance bands from sources of dust emissions; and Ground truthing of potential receptors through fieldwork and by the ArcelorMittal Liberia Limited Social Development Team, to confirm the number, nature and level of use. Since the production of the Phase 1 ESIA, further visits to the mine site, railway corridor, port and surrounding area have been made by the URS Phase 2 ESIA project team. Recent information gained during the site visits has been combined with the existing Phase 1 information. Sensitive Flora and Crops Sensitive flora and crops were identified through: a consideration of areas where dust deposition could occur at a rate sufficient to cause stomatal blocking and a consideration of chemical analyses of soil and ore samples which was also made to determine whether they contained any toxic compounds or elements. Where either of these factors was identified, the location and extent of any areas potentially affected by either effect were further examined to determine whether they supported any sensitive crops or flora Baseline Dust and Particulate Matter Prior to the commencement of work on Phase 1 of the project, there was no historical recorded data relating to dust and particulate matter concentrations within the study area. A programme of dust monitoring was therefore undertaken to provide a baseline against which any future operational impacts can be compared. Monitoring stations measuring concentrations of airborne particulate matter, dust deposition rates, and wind speed and direction were installed at four sites selected to represent locations where the larger population centres are closest to areas of mining, railway and port activity, i.e. where the potential for effects is likely to be greatest, so that they would be representative of worst case exposure scenarios where the main populations are located. The four monitoring (GPS derived) locations were as follows: Gbapa within Gbapa village (Grid Ref: longitude, latitude: , ). Chosen since Gbapa is the closest major community to the Gangra-Yuelliton ore stockpile/loading area at km 250; Zolowee the monitoring station is located within Zolowee village (Grid Ref: longitude, latitude: , ). The closest major community to the Tokadeh ore stockpile/loading area; 12

14 Sanniquellie within the grounds of the Superintendent s building (Grid Ref: longitude, latitude: , ). Sanniquellie is the closest major community to the rail line; Buchanan within the ArcelorMittal Liberia Limited compound area, approximately 500 metres from the boundary with the local community, but close to the proposed stockpile areas (Grid Ref: longitude, latitude: , ). At each site the following monitoring equipment was installed: Osiris monitors for the continuous measurement of particulate matter concentrations. The instruments operate on the principle of light-scattering and provide a continuous and simultaneous measurement of particles in the PM10 and PM2.5 fractions3. These data provide an indication of both airborne dust concentrations and particle size and are therefore relevant to consideration of impacts on human respiratory health. Frisbee dust deposit gauges for the measurement of dust deposition rates. Samples are collected over a four week period and then returned to the laboratory for the subsequent determination of total insoluble solids. These data provide a measure of dust deposition rates and are therefore relevant to consideration of potential effects upon disturbance to humans in terms of amenity effects, and effects upon crops and sensitive ecosystems. Data collection to inform the Phase 1 ESIA was undertaken between November 2008 and May The equipment did continue to collect data beyond this date as part of an ongoing programme, but site specific difficulties have meant that it has not been possible to obtain more recent periods of monitoring data with acceptable capture rates. Baseline dust and particulate matter data for this assessment has therefore been taken from the same period as that presented in the Phase 1 ESIA Survey of Baseline Nitrogen Dioxide and Sulphur Dioxide Emissions from the power plants to be constructed in association with the Phase 2 development of the Tokadeh mine site and Port at Buchanan would have the potential to increase local concentrations of NO 2 and SO 2. In order to evaluate baseline concentrations of these two pollutants, a project specific passive diffusion tube monitoring survey has commenced in the area around the development sites. The sampling is programmed to run for six months and was initiated during March The survey will therefore capture data in both the wet and dry seasons and, once complete, will be representative of annual mean baseline conditions. Passive diffusive samplers are simple devices which are widely used for the measurement of ambient NO 2 and SO 2 concentrations. Whilst they do not offer the same precision or accuracy as automatic chemiluminescent monitoring, their robust nature and ease of use make them a useful tool to provide evaluation of long term pollutant concentrations. The samplers consist of a 7 cm long acrylic tube that can be sealed at both ends. The tubes are normally left in place for one month, after which they are replaced by another tube, sealed and returned to a laboratory for analysis. One end of the device contains stainless steel grids coated with a substance that adsorbs SO 2 or NO 2, which is then subsequently analysed to determine the mass of the pollutant captured by the tube. Exposure commences when the inlet cap is removed, which sets up a concentration gradient within the tube, such that molecular diffusion occurs towards the coated grid. The diffusion tubes were placed at seven locations in the area around Tokadeh and three locations within Buchanan. Some locations were selected to be representative of background conditions in rural and residential areas, while others were chosen to be representative of human exposure in locations next to existing sources of emissions such as roads and power generation equipment. The sites selected for the survey are listed in Table PM 10 and PM 2.5 is fine particulate matter with an aerodynamic diameter of less than 10 and 2.5 micrometres respectively. 13

15 TABLE 3.1: DIFFUSION TUBE MONITORING LOCATIONS FOR NO 2 AND SO 2 Site ID Location Descriptor Grid Reference T1 Yekepa Workshop, adjacent to existing power plant , T2 Yuelliton Gangra Road, rural background , T3 Gbapa, adjacent to main road , T4 Tokadeh Magazine, within mine site , T5 Zolowee, close to main road , T6 Makinto , T7 Geh , B1 AML office at Buchanan , B2 Moore Town Gate, Buchanan , B3 North of Buchanan Port Gate , Meteorological Conditions and Topography In order to inform the prediction of air quality it is necessary to determine the dispersion characteristics in the vicinity of emission sources. In addition to the particle size distribution, these are also influenced by wind direction, rainfall and topography. Furthermore, hourly sequential meteorological data is required for use in dispersion modelling studies, and it is important that the data used are representative of conditions at the point of emission and the surrounding modelled domain. Data generated from three weather stations installed for this ESIA in Nimba and Buchanan were considered for use in this study. The three meteorological stations are as follows: Within the workshop compound at Yekepa (Grid Ref: , ); Tokadeh Magazine within the mine site (Grid Ref: , ); and Buchanan (Grid Ref: , ). Topographical details were established through a review of topographical maps (generated through Light Detecting and Ranging (LiDAR) techniques) and field visits, while vegetation was assessed through field visits. 3.2 Prediction and Evaluation of Air Quality Effects Emissions of Dust and Particulate Matter Dust emissions, dispersion patterns and impacts are difficult to predict due to the lack of reliable emissions factors for the wide range of Scheme activities that might give rise to dust. Emission factors can be derived from a number of sources including documents published by the United States Environmental Protection Agency (USEPA, 1998). However, The USEPA emissions factors predominantly relate to US Western surface coal mining and there is significant uncertainty regarding many of these factors, particularly when they are applied to mining operations that differ to those from which the source data were collected. Furthermore, it is impossible to check (or verify) the performance of dispersion models used for these types of studies and at best the predictions can only provide a crude estimate of the likely dispersion patterns that might arise. A further consideration is that dispersion models 14

16 are unable to accurately predict dust concentrations during low wind speeds or calms (wind speeds below 0.5 m/s). This is because dust transport and dispersion is principally governed in the model by the wind speed and direction. The approach adopted to predict and evaluate the effects of dust emissions for the ESIA has therefore been based on the fact that dust concentrations and deposition rates decline rapidly with increasing distance from the emission source, particularly during calm conditions. This is due to deposition of the larger particles, and the dilution and dispersion of fine particles. Consideration has thus been given to the various Scheme activities with regard to the location of sensitive receptors (Section above), taking into account the: Level of likely dust generation based upon the type, location, duration and frequency of activities; and Size of the dust particles, and hence the likely distance they could travel; Evidence has been drawn from published reports and other relevant guidance, with regard to the distances over which significant impacts might occur. Particle Size The term dust is commonly taken to represent particles between 1 and 75µm diameter. Particles that are less than or equal to 10µm diameter are commonly referred to as PM 10, those less than or equal to 2.5µm diameter are referred to as PM 2.5. The larger particles are associated with public perception of dust nuisance and the deposition of dust onto vegetation. The smaller particle sizes, i.e. the PM 10 and PM 2.5 fractions are most strongly associated with human health effects including respiratory and other diseases. The larger dust particles (those above 30µm) make up the greatest proportion of dust emitted from mine workings, and tend to fall out of the atmosphere rapidly, rarely travelling beyond about 100 metres from the emissions source (ODPM, 2005; Department of the Environment Minerals Division, 1995). The intermediate size particles (10-30µm) are likely to travel up to metres. The fine particles (less then 10µm) are deposited very slowly and can travel distances of 1000 metres or more. However, the latter comprise only a small proportion of dust emitted from mine operations and this, combined with the effects of dispersion and dilution, means that concentrations are unlikely to be significant beyond about 200 metres of the emission source. Dust emissions can arise from both mechanical suspension or from wind blow. The former refers to site activities such as ore extraction, materials handling, or the passage of vehicles across haul routes, which cause dust particles to be entrained into the atmosphere. The latter refers to the wind lifting surface-bound dust into the atmosphere directly: this is a function of wind speed and turbulence, but dust particles (below about 75µm) are resistant to these erosive forces until wind speeds increase above about 6 metres/second Dispersion Modelling of Power Plant Emissions Overview This section describes the model used to evaluate the impact of emissions to air on ambient pollutant concentrations in the vicinity of the concentrator power plant site. Dispersion Model Selection The air quality impacts of the proposed power plant are best evaluated using a refined, near field (less than 50 km from the emission source) Gaussian Plume Dispersion Model, which is able to calculate maximum ground level concentrations at receptors close to the plant stacks. Gaussian models assume that pollutants do not decompose in the atmosphere, and therefore do not account for the long-range transport of atmospherically reactive pollutants. They are designed to produce results that are close to monitored values. 15

17 The assessment of emissions from the main stacks has been undertaken using ADMS 4.2, supplied by Cambridge Environmental Research Consultants Limited (CERC). ADMS is a modern dispersion model that has an extensive published validation history (CERC, 2007). In particular, this version of ADMS has proven benefits for modelling meteorological conditions where there is a high incidence of calm conditions. Modelled Scenarios The dispersion modelling has been undertaken for a single operational scenario. The power plant would contain ten engines, but the power demands of the concentrator are very stable and the base load would consist of nine engines running at any one time, with one in standby (N+1 configuration). The assessment has therefore considered emissions from nine engines running continuously without any further emission variation. Due to the current uncertainty surrounding the guaranteed sulphur content of fuel available for delivery to Buchanan, the assessment has considered two scenarios for SO 2 emissions; one scenario in which the plant would be operated according to IFC guidelines with maximum 2% sulphur fuel, and one in which the plant would be operated with 4% sulphur fuel (the current guaranteed level given by the fuel supplier). Dispersion Model Inputs The general model conditions used in the assessment are summarised in Table 3.2. Other more detailed data used to model the dispersion of emissions is considered below. TABLE 3.2: GENERAL ADMS MODEL CONDITIONS Variable Surface Roughness at source Receptors Receptor Location Source location Emissions Sources Meteorological data Terrain data Buildings that may cause building downwash effects Input 1.0 m Selected discrete receptors x,y co-ordinates determined by GIS, z = 1.5 m x,y co-ordinates determined by GIS Data provided by AMEC project team 9 main exhaust stacks in 3 clusters Hourly sequential dataset from ArcelorMittal station at Tokadeh Magazine (period 1/07/2011 to 30/06/2012) Complex terrain, 64x64 ADMS terrain file derived from LIDAR data Turbine Hall Emissions Data The physical properties of the combustion emission sources, as represented within the model, are presented in Tables 3.3 and 3.4. This data has been taken from the scheme design drawings for the concentrator and emissions information provided by the AMEC project team and the engine manufacturers, Caterpillar. 16

18 TABLE 3.3: MODELLED EMISSION SOURCES Release Point Grid Reference 16CM32_ , CM32_ , CM32_ , CM32_ , CM32_ , CM32_ , CM32_ , CM32_ , CM32_ , Note: The x and y coordinates listed are referenced to the WGS co-ordinate system, UTM zone 29N TABLE 3.4: PHYSICAL PROPERTIES, COMBUSTION PLANT STACKS (PER ENGINE) Parameter Unit Value Stack Height m 32 Effective internal stack diameter m 1.0 Flue temperature C 315 Volume flux m 3 /s 27.0 Stack gas velocity m/s 34.3 NO X mass emission rate g/s PM 10 mass emission rate g/s 0.84 SO 2 mass emission rate (2% S) g/s SO 2 mass emission rate (4% S) g/s CO mass emission rate g/s 1.34 The modelled pollutant emission rates (in g s -1 ) are determined by the emissions limits set out within Table 6(A) of the IFC guidance for Thermal Power Plants (Emissions Guidelines for Reciprocating Engines of bore size less than 400 mm, Liquid Fuels, Plant >50 MWth to <300 MWth). These limits are: Particulate Matter: 50 mg/m3 Sulphur Dioxide (SO2): 1,170 mg/m3 or <2% S in fuel Oxides of Nitrogen (NO2): 1,460 mg/m3 Carbon Monoxide (CO): 70 mg/m3 Pollutant mass emission rates from the power plant engines (in g/s) have been calculated by multiplying the IFC emission limit concentrations by the volumetric flow rate at full load operation (at reference conditions of 15% O 2, dry). For the 4% fuel sulphur scenario, the mass emission rate has been calculated using an indicative emission concentration supplied by Caterpillar, of 2500 mg/m 3. 17

19 Modelled Domain Ground-level concentrations of the modelled pollutants have been predicted at 7 discrete air quality sensitive receptors, as listed in Table 3.5. The receptors have been selected as locations where the nearest residential properties are found in each direction from the power plant site. The receptor locations have been selected to be representative of residential properties in the area around the site, and can be considered to be representative of other receptors in their vicinity. The height of the receptors has been set at 1.5 m. TABLE 3.5: MODELLED DOMAIN, SELECTED DISCRETE RECEPTORS Receptor Description Grid Reference R1 Zolowee N , R2 Zolowee S , R3 Makinto , R4 Geh , R5 Gbapa , R6 Vayanpa , R7 Kpadapa , Note: The x and y coordinates listed are referenced to the WGS co-ordinate system, UTM zone 29N Meteorological Data Hourly sequential meteorological data is required for use in dispersion modelling studies, and it is important that the data used are representative of conditions at the point of emission and the surrounding modelled domain. In many situations, data collected from the nearest airport can be considered appropriate, but in this case the nearest such available sources are from Roberts International Airport, Monrovia (approx 270 km to the SW) and Abidjan, Ivory Coast (approx 560 km SE). Both of these sites are coastal in nature, are some distance away, and cannot be considered to represent climatic conditions in the area around the mine site at Tokadeh. For this reason, it is necessary to use project specific meteorological data with the model. The climate in the Nimba Range area of Liberia is unique, and meteorological monitoring and analysis for the project has shown that there are relatively large variations in climate over even short distances. For example, the prevailing wind conditions at Tokadeh are very different from observations made at Yekepa. For this reason, data collected at the mine site itself has been selected for use in the assessment. The measurement site is around 1 km to the west of the proposed power plant. The station was set up and began collecting data in July This assessment uses data collected between 1 July 2011 and 30 June 2012, a period of twelve months. The data capture rates have been good over the majority of the year, although technical issues meant that data was not collected between 16 March and 24 April Despite this limitation, the dataset will still have captured the seasonal variation in climatic conditions and prevailing wind direction at the site during the monitoring period. Although multiple years of meteorological data would normally be employed in a study such as this one, it is considered that the use of local data is of higher importance in this case, rather than a longer period from a location that is less representative of local climatic conditions at the mine site. 18

20 A visual representation of the wind speed and direction data used in the assessment is shown in the wind rose presented in Figure 3.1. The assessment does not use the wind roses to infer the magnitude or frequency of impacts at any receptor. Instead, the hourly sequential observation data are used in the dispersion model to calculate robust estimates of impacts. Figure 3.1: Wind Rose for Tokadeh Magazine, 1 July 2011 to 30 June 2012 Building Downwash Effects The buildings that make up the power plant have the potential to affect the dispersion of emissions from the main stack. The ADMS buildings effect module has therefore been used to incorporate building downwash effects as part of the modelling procedure. Buildings greater than approximately one third of the preferred stack height have been included within the modelling assessment. At the current time, the design of the power plant buildings and fuel tanks have not been finalised, so the evaluation of building downwash has been based on preliminary building dimensions. Only the turbine hall is around one third of the proposed stack height. The HFO tank design has not been finalised and so the effect of the HFO tanks on the plume has not been included. The proposed location of the tanks is at least 50 m from the nearest stack, and so the effect of the tanks on the plume is unlikely to be significant unless they are well in excess of 10 m in height. 19

21 The building dimensions, as represented within the model, are presented in Table 3.6. The turbine house has a sloping roof, and has been represented within the model as being the distance from the floor to the apex (not including the roof vent). TABLE 3.6: BUILDING PARAMETERS Building Grid Reference of Centre Point Length (m) Width (m) Height (m) Angle ( ) Turbine Hall , Note: The x and y coordinates listed are referenced to the WGS co-ordinate system, UTM zone 29N Terrain The land in the area around the concentrator site is complex, with sharp changes in gradient and level within the modelled domain. As the terrain would influence the flow of air over and around the land between the emission point and sensitive receptors, the consideration of terrain effects within the model is necessary. The terrain grid has been prepared from the outputs of a 1 m resolution LIDAR survey of the mine site and surrounding area. GIS software was used to generate the 64 x 64 node terrain file for ADMS. Although the final ground levels within immediate vicinity of the power plant site would change slightly during work to construct the concentrator and power plant, the differences would not be great enough to significantly affect the results of the modelling and no change has been made to the terrain file to place the base of the stack in relation to the surrounding terrain. Surface Roughness A surface roughness of 1.0 m was used within ADMS. This option is considered as representative of the predominantly rainforest landscape within the study area. As the meteorological data used was taken at the site itself, the same surface roughness value was used for the meteorological station. NO X to NO 2 Conversion Emissions of NO x from the main stack will consist mainly of nitric oxide (NO) at the point of release, oxidising within the atmosphere to form NO 2 as it moves downwind. At the point of release, approximately 95% of the NO X emission will be in the form of NO, with the other 5% consisting of NO 2. In the assessment of the impact on annual mean concentrations of NO 2, a worst case approach has been taken and a 100% NO x to NO 2 conversion rate at ground level has been assumed in the calculation of long-term annual mean calculations. With regard to short-term impacts, even within such a rural environment with relatively high ozone (O 3 ) concentrations, the oxidisation of NO to NO 2 is a relatively slow process, and for this reason it is considered that the assumption of 100% conversion within the modelled domain is excessively conservative in locations close to the power plant stacks. For this reason, the NO X to NO 2 conversion rate has been determined by the ADMS chemistry module in the calculation of short-term hourly concentrations. The ratio of NO : NO 2 in the plant stacks was assumed to be 95% : 5% and the ambient O 3 concentration in the atmosphere around the facility was assumed to be a constant 50 µg/m 3. The O 3 concentration used in the assessment has not been derived from project specific sampling, but has been taken from sampling carried out in Cameroon. The sampling was undertaken outside of Douala, in a rural area within the tropical zone, and is considered to be broadly representative of the area around Tokadeh. This value used is also at the high end of the range reported in the literature for tropical latitudes (Folkin et al, 2002). It is conservative to use a higher background O 3 concentration, which would result in higher conversion rates from NO to NO 2. Specialised Model Treatments Emissions have been modelled such that they are not subject to dry and wet deposition. This assumption of continuity of mass is likely to result in an over-estimation of impacts at receptors. 20

22 4 BASELINE CONDITIONS 4.1 Receptor Locations Mine Area Prior to the resumption of mining activities, a number of hamlets were identified within the area around the mine sites that were in a semi-abandoned state, but were potentially occupied for some of the time. Residential properties within the area potentially affected by dust from mine operations were identified. No schools or hospitals were found. During Phase 1 of the project, a resettlement programme was put in place to relocate residents of properties and users of farmland situated within a 500 m exclusion zone around the boundary of the mine sites. The Phase 1 resettlement programme was completed. The locations of the main settlements as well as hamlets and individual properties used by communities and any particularly sensitive receptors (e.g. hospitals and schools) within the specified distance bands are listed in Appendix B with a description of their nature and level of use, and shown on Figures 8.1 and 8.2 of the Draft Final Report for Environmental Impact Assessment Phase 1 (Atkins 2009) Railway Corridor The railway and service road pass close to a number of important settlements in three counties, notably: numerous large settlements and towns including Compound 3, Isaac Anjuah and St. John in Grand Bassa; Botota, Zowenta and Duo Town in Bong County; Gbeden, Sanniquellie and Sekyi Kempa in Nimba (Figure 8.3 of Draft Final Report for Environmental Impact Assessment Phase 1 (Atkins 2009)). Between these settlements are a number of other towns and smaller settlements which could be affected by increased dust levels due to payload activities and vehicle movements along the service road. Population distribution is most dense in Grand Bassa and sparsest in Nimba County. The Sanniquellie hospital is located close to the railway line (50m at the closest point) Greenhill Quarry There are a number of residential properties located near to Greenhill Quarry, which was the source of rock used in the rehabilitation of the railway and other parts of the scheme Port (Landside) The town of Buchanan lies to the north of the port concession area. There are no residential properties within 200 m of the proposed unloading and stocking area. 4.2 Dust and Particulate Matter Sources of Baseline Dust and Particulate Matter Due to the tropical climate in Liberia and the predominantly agricultural activities carried out on the land, most of the ground in rural areas is permanently covered by lush vegetation and rain forest. For this reason, even during the dry season there are few naturally occurring sources of windblown dust. The principal local sources of fugitive dust would therefore be from: vehicles (in particular heavy lorries) travelling on the local roads, which are largely unpaved; emissions from domestic cooking and farming. During dry conditions, vehicle movements across unsurfaced roads can raise substantial dust plumes in their wake, which may affect both populations and vegetation close to the roads. Emissions from domestic cooking and farming may represent localised sources of particulate matter, particularly when vegetation is burned by local people to clear land for the planting of 21

23 crops. Most farming in the area around the mine sites is, however, carried out without the use of machinery and so dust emissions produced from the working of the land are generally small in scale. Larger dust particles are deposited within a relatively short distance of the emission source, the process of which is accelerated by the presence of dense vegetation at the sides of roads and the borders of working areas. During the field visits undertaken during the dry season in March 2012, heavy dust staining could be observed on vegetation immediately adjacent to unsurfaced roads, but at a distance of 10 m to 20 m from the roadside, the evidence of dust deposition was greatly reduced. It is therefore likely that in areas where vegetation screening is in place, the majority of re-suspended or windblown dust from roads and other uncovered areas of ground is deposited within 50 m to 100 m of the source of the emission. Finer dust particles may be transported much greater distances depending upon the weather conditions. The amenity of residents of villages situated in close proximity to the main traffic routes is particularly affected by dust from passing traffic during dry conditions. Larger vehicles, and vehicles travelling at higher speeds, produce much larger emissions of fugitive dust. Measures to reduce vehicle speeds, such as speed humps, or the wetting of the road surface, were observed during the field visit to make a large positive impact on the magnitude of the emission. Similarly, in areas where roads are sealed, there is a marked decrease in dust emissions to the point where there is no significant impact on amenity Measured Dust and Particulate Matter Concentrations A summary of the PM 10 and PM 2.5 concentrations measured over the period November 2008 to May 2010 is provided in Table 4.1. The seasonal variability of PM 10 concentrations during this period is reported in greater detail in Table 4.2. Monitoring at Gbapa ceased in April TABLE 4.1: MEASURED PM 10 AND PM 2.5 CONCENTRATIONS Site Data Capture Period Mean PM 10 PM 2.5 Gbapa 56% Zolowee 43% Sanniquellie 60% Buchanan 54%

24 TABLE 4.2: MONTHLY VARIATION IN MEASURED PM 10 CONCENTRATIONS, NOV 2008 TO OCT 2009 Season Month Buchanan Sanniquellie Zolowee Gbapa Mean Std Dev. Mean Std Dev. Mean Std Dev. Mean Std Dev. Dry Nov Dry Dec Dry Jan Dry Feb Dry Mar Dry Apr Wet May Wet Jun Wet Jul Wet Aug Wet Sep Wet Oct The following conclusions may be drawn from these data through comparison with the guideline standards outlined in Table 2.1: The period mean PM 10 baseline concentrations exceed the WHO Interim Target 1 annual mean guideline at Gbapa and Zolowee. Period mean PM 10 concentrations at Sanniquellie and Buchanan are within the assessment criteria of 70 µg/m 3 ; There is considerable variation about the mean for PM 10 concentrations at all sites with greater variations occurring in drier conditions when winds are most likely to transport particulate matter into the study area from the Sahara, increasing the likelihood of the 24- hour mean concentrations becoming elevated; and The period mean PM 2.5 concentrations are well within the WHO Interim Target 1 annual mean guideline at Gbapa, Sanniquellie and Buchanan. At Zolowee, the period mean PM 2.5 is very close to, but within the criteria. The data indicates that there is a risk of exceedance of the annual mean PM 2.5 guideline in the area around the monitoring site. Considerable caution needs to be applied to the above interpretation of these observations as data capture rates at all locations were low due to site-specific difficulties. Further analysis of the data shows that there is a marked seasonal effect on airborne concentrations of particulate matter. Measured concentrations during the wet season are typically reduced as dust emissions from unpaved roads would be substantially suppressed, although episodes of high particulate matter concentrations have been recorded in all months. The dry season also coincides with the Harmattan winds, which can cause regional-scale dust episodes lasting for extended periods. Although the Harmattan is likely to have affected measured particulate matter concentrations during the monitoring period, local sources are still considered to be the predominant source of emission. The monitoring period at Gbapa was shorter than at the other sites, data collection only took place between November 2008 and April The period mean and 24-hour mean concentrations reported for this site are therefore very likely to have been elevated by the 23

25 monitoring period coinciding with the dry season and the Harmattan. Monitoring over a longer period would probably have given period mean data lower than that shown in Table 4.1. The proximity of the Zolowee monitoring site to the unpaved main road from Sanniquellie to Yekepa, means that recorded levels are inevitably dominated by traffic-related dust and this is reflected in the less obvious seasonal pattern at this site. The levels at Buchanan may also have been influenced by the close proximity to a gravel road. As many residential properties within the study area are located a similar distance from such roads, the baseline data can however be considered to be representative of baseline conditions at sensitive receptors. Long-term concentrations of particulate matter within the wider area would be likely to be lower than the values reported. The measured PM 2.5 concentrations are significantly lower than the PM 10 levels (ratios ranging from 0.13 at Zolowee to 0.19 at Sanniquellie). This indicates a significant coarse dust emissions source, consistent with re-suspended dust from unpaved roads Measured Baseline Dust Deposition Rates A summary of the baseline dust deposition rates measured between November 2008 and January 2009 at Gbapa, Zolowee, Sanniquellie and Buchanan, is provided in Table 4.3. TABLE 4.3: MEASURED DUST DEPOSITION RATES Site Monitoring Period Dust Deposition (mg/m 2 /day) Gbapa Zolowee Sanniquellie Buchanan 20/11/08 22/12/ /12/08 28/01/ /11/08 22/12/ /12/08 28/01/ /11/08 22/12/ /12/08 28/01/ /12/08 06/01/ /01/09 28/01/ The dust deposition rates at all four sites are within the South African standard for dust deposition within residential areas (Table 2.1). The highest levels are seen to have occurred at Zolowee and Buchanan, while the levels recorded at Gbapa and Sanniquellie are much lower. As in the case of measured particulate matter concentrations (Table 4.1), it is highly likely that the higher values reported for the Zolowee and Buchanan stations are attributable to the proximity of unpaved traffic routes. While this can be considered to be representative of conditions during the dry season near to such roads, baseline dust deposition within the wider area would be lower. Considerable caution should be applied to interpretation of these data, as they represent only two month s sampling during the dry season Measured Baseline Nitrogen Dioxide and Sulphur Dioxide The baseline diffusion tube survey in the area around the Tokadeh mine site and Buchanan commenced in March The survey was programmed to run for six months, but site specific difficulties have resulted in some loss of data. At the time of writing, the results of monitoring from March to May 2012 were available for NO 2, and March to July 2012 in respect of SO 2. These results are presented in Table

26 Although concentrations measured from month to month will vary due to meteorological conditions and seasonality the results indicate that, as expected, baseline concentrations of combustion pollutants within the air quality study area are very low in most locations. Slightly elevated concentrations, in comparison with background levels, are found in close proximity to roads and other emission sources. In Buchanan, slightly higher concentrations of NO 2 were recorded than in the area around the mine sites. This is predominantly due to a greater level of road traffic movements in the port area in comparison to the mine site, emissions from ships at berth, loading shovels and rail locomotives. The greatest NO 2 value was recorded in close proximity to the train unloading and stocking facility at the port. At all the monitoring locations in the two areas surveyed, the data indicates that there is no existing risk of a breach of the long term and short term air quality standards for NO 2 and SO 2. TABLE 4.4: DIFFUSION TUBE MONITORING RESULTS, MARCH TO JULY 2012 Mean Measured Pollutant Concentration (µg/m 3 ) Site ID Location NO 2 SO 2 T1 Yekepa Workshop T2 Yuelliton Gangra Road T3 Gbapa T4 Tokadeh Magazine T5 Zolowee T6 Makinto T7 Geh B1 AML office B2 Moore Town Gate B3 Port Gate Greenhouse Gas Levels Total emissions of carbon dioxide from Liberia in 2005 were estimated at 358,000 tonnes (CDIAC, 2008), principally arising from liquid fuel combustion. Emissions have remained largely unchanged since 1990, and represent only a very small proportion of the total emissions arising from the Sub-Saharan African region. 4.4 Meteorology A detailed analysis of meteorological conditions in the Northern Nimba area of Liberia was undertaken as part of Phase 1 of the project by the UK Meteorological Office (Met Office, 2010), using available data from long term monitoring and project specific sources. The primary aims of the study were to identify the potential for extreme wind speed and rainfall events to occur, but the report also summarises the overall climatic conditions in the vicinity of the mine sites, concentrator and rail loading facility. The climate of Liberia is equatorial and is characterised by hot, humid conditions for most of the year. There is a distinct, seasonal pattern of rainfall, with heavy rain between April and October, and a relatively drier period extending from November to March. 25

27 Due to the high rainfall during the wet season, the potential for significant dust emissions during this period will be substantially reduced. The potential for significant dust emissions is therefore likely to be restricted to the relatively short, dry season. In the Yekepa area, winds are predominantly from the south to south-westerly directions, and are characterised by very low speeds, generally below 1.5 metres/second, and with a very high frequency of calms (<0.5 metres/second). At Tokadeh, the winds are predominantly from the north-west, and are again characterised by very low speeds and high frequency of calms. This disparity between the Yekepa and Tokadeh sites may be related to local topographic effects, causing channelling of the winds at ground level. However, caution must be applied to any detailed interpretation of these data, as measurements were only carried out for a short period, and a substantial number of hours recorded wind speeds below 0.5 metres/sec. At low wind speeds, there is potential for the anemometer to stall until the wind speed exceeds the inertia threshold for the instrument. This part of West Africa is also influenced by dry and dusty trade winds which blow from the Sahara to the Gulf of Guinea during the winter period, between the end of November and the middle of March. Known as the Harmattan, these winds may carry a significant amount of dust into the atmosphere causing local hazes that can limit visibility for several days at a time. Within this part of West Africa, up to 50kg/ha of dust can be deposited during the Harmattan (Ogunstein, 2007). However, such episodes will be regional in scale, and are not directly relevant to this assessment, as local sources of dust are considered to be more important. Low wind speeds will significantly reduce the dispersion of dusts and the distances over which dust particles will travel, thus limiting the potential for significant effects to well within the distance bands identified in Section 1.2. The low wind speeds will also significantly reduce the potential for wind blown resuspension. 4.5 Topography and Vegetation The topography in the vicinity of the mine area is dominated by the range of mountains comprising Tokadeh, Beeton, Gangra and Yuelliton to the west, with peaks rising to 1040m, and the Nimba Ridge rising to 1500m to the east. To either side of the Nimba Ridge, the land falls away to a series of lower hills and into a plateau of broad plains and small hills, between about m elevation. The extraction areas at Tokadeh and Gangra-Yuelliton lie between 600 and 1000m elevation with the closest towns of Zolowee and Gbapa at approximately 400 metres elevation. South of Mount Tokadeh, the landscape becomes more open and undulating, and is less influenced by the high ground of the mountain ranges. This area includes Sanniquellie, which forms one of the largest settlements to the south of the mining area. Further south from Sanniquellie, and following the line of the railway and access road, the topography becomes flatter with a rolling landscape as it broadly follows the path of the Saint John River. At Buchanan, the topography is generally flat, with the port area lying to the south of the town. The elevated locations of the extraction areas with respect to Zolowee and Gbapa will serve to increase the dilution and dispersion of dust emissions from the mining operations, and will reduce the potential for significant effects to occur. The topographical features at other locations are unlikely to be important in terms of air quality effects. The dominant vegetation cover at Gangra-Yuelliton is dense forest, but with localised areas of wetland or forest clearance and agriculture. Mount Tokadeh is also largely forested, but with some areas subject to extensive clearance resulting from the former mining activity and associated infrastructure. The villages of Gbapa and Zolowee lie within relatively open areas, formed from the predominant use of slash and burn farming activities. The predominant crops are pineapple, cassava and rice. The settlements are generally screened by belts of vegetation that divide the houses from the adjacent land uses, although some buildings are often located within close proximity to the main road. 26

28 Most of the railway corridor is dominated by agricultural land with small forest. The area identified for the port redevelopment is largely hard standing with some patch scrub. Much of the benefit of the attenuation of dust effects is gained through the offset provided between the operational areas and sensitive receptors. However, the dense mixed vegetation in the vicinity of the mine area will also serve to act as a screen, removing dust from the atmosphere and further reducing the potential for significant effects to occur. The belts of vegetation surrounding the settlements will also serve to reduce dust effects, albeit to a lesser extent. The vegetation at other locations is unlikely to be important in terms of air quality effects. 4.6 Ore and Surface Soil Characteristics Ore analysis data from samples provided by ArcelorMittal Liberia Limited indicate a moisture content of around 8% for DSO. The moisture content of ore leaving the concentrator will be around 6%. These values lie within the reported range of material moisture contents, e.g. between about 0.5 to 22% (USEPA, 1998), and will serve to reduce the potential for dust emissions. An analysis of particle size distribution carried out on 63 soil samples collected by WSAI from the Tokadeh area in 2006 revealed an average silt content (in this case taken as the <63µm fraction 4 ) of 18.5%. An analysis of particle size distribution has also been carried out on an ore sample to inform this ESIA Report. The results are summarised below: Less than 75µm 16.5% Less than 30µm 14.7% Less than 10µm 12.1% (82% of less than 30µm fraction) Less than 2.5µm 9.1% (62% of less than 30µm fraction) The result for the larger size fraction is reasonably consistent with that reported from the 2006 soil samples. The USEPA has published generalised particle size distributions for a range of materials, including materials handling and processing of aggregate and unprocessed ores (USEPA, 1996); these data are reported as the cumulative fraction of the Total Suspended Particulate (i.e. the less than 30µm fraction). These generalised data indicate that the less than 10µm fraction represents about 51% (of the less than 30µm fraction), whilst the less than 2.5µm fraction represents about 15%. In comparison to these generalised values, the less than 10µm and 2.5µm fractions of the DSO ore are relatively high, but it is not known to what extent the particles will tend to aggregate when not mechanically disturbed. 4 This is broadly consistent with the definition of silt (<75µm) 27

29 5 OPERATIONAL PHASE IMPACTS The phasing of the Scheme and the scheduling of associated activities is set out in detail within the main text of this ESIA. These details are not repeated here, but it is useful to highlight some of the important assumptions that have informed this air quality assessment: The need to strip topsoil or other overburden is expected to be minimal within the mining areas; The ore has a relatively high moisture content and the fraction of fine material (less than 10µm) is high (see Section 4.6). Moisture does, however, have a proportionally greater beneficial effect on smaller particles, making them less likely to become airborne; The facility will be available to spray the haul routes at Tokadeh and Gangra-Yuelliton with water during dry weather conditions; The concentrator power plant would operate on a continuous basis at base load, and due to the stable operating conditions it is assumed there is no variation in emissions over time; and Ore wagons will not be covered. The consideration of baseline conditions has identified that airborne particulate matter concentrations can be higher than the WHO interim target values set out in Table 2.1, particularly so at sensitive receptor locations in close proximity to unsealed roads. Also, during periods when the Harmattan wind brings large amounts of suspended particulates into the area, overall background levels can become elevated for days or even weeks at a time. The primary focus of this prediction and evaluation of effects is therefore the expected incremental change in conditions due to the Scheme, rather than direct reference to absolute concentrations mostly composed of material released from sources present in the baseline scenario. 5.1 Fugitive Emissions of Dust and Particulate Matter from the Mine Area Sources of Dust The principal sources of dust emissions will be associated with a range of Scheme activities including: Initial site preparation and construction works; Soil stripping/overburden handling and storage; Drilling and blasting of ore within the extraction area, where this is necessary; Excavation of ore material; Transportation of ore from the mine to the concentrator processing area; Ore crushing and blending; Screening and stockpiling activities; Processing of ore within the concentration plant; and Loading of ore onto rail wagons for onward transport Prediction of Effects on Human Populations The closest main population settlements to these dust generating activities are the towns of Gbapa and Zolowee (1km and 2km respectively). These are too distant to be affected by either dust nuisance, associated with the larger particles, or health effects associated with fine particulate matter. A resettlement programme prior to the implementation of Phase 1 relocated 28

30 the residents or users of other scattered buildings and farming land within the spatial scope that could potentially be affected by dust impacts. For this reason, there are a very limited number of sensitive receptors within the areas that could be affected by the mining, ore processing and loading operations. The following conclusions on the operational activities to be undertaken can be drawn: The greatest potential for fugitive dust emissions is likely to arise from the passage of trucks along the haul routes and access roads. This is highly unlikely to represent a problem during the wet months, when there is consistently more than 1mm of rainfall per day. In the absence of suitable mitigation measures there could be the potential for significant impacts on human populations to occur during periods of dry weather, particularly at those remaining properties within 200 metres of the Gangra-Yuelliton haul road. Any effects would be occasional, temporary and minor in nature, given the locations of the closest receptors and the high incidence of low wind speeds recorded in the area. Occasional dust plumes extending beyond the boundary of the mine footprint cannot be completely excluded, and will be dependent upon local weather conditions. Whilst the smaller dust particles can potentially be transported up to one kilometre, the dispersion will be limited by the very low wind speeds recorded in this area, and it is expected that any impacts would mostly be restricted to within 200 metres of the mine footprint boundary. There are no sensitive receptors within 200m of the Gangra Yuelliton mine area or the Tokadeh footprint. Without mitigation, particularly during dry periods of weather, or during periods of activity close to the mine footprint boundary, it is potentially possible that there could be significant impacts on human receptors. The proposed primary crusher is an enclosed process, which along with the inherent moisture content of the run of mine ore would limit the potential for fugitive dust emissions. Dust collection and suppression systems are to be provided for each crusher, comprising a misting system on the dump pockets, plus bag filters and fans at the crusher discharge point to the apron feeders and transfer to the conveyors. These measures, combined with a distance of more than 2 km to the nearest residential properties, mean that no significant impacts on human receptors are expected. Crushed ore would be transferred to the blending and stocking area by belt conveyor, thereby removing the dust generation potential associated with truck movements along an unsurfaced haul route. Crushed ore blending system and stockpiling / reclaiming equipment would be fitted with water cannons for dust suppression. The stockpiles are at least 1.5 km from the nearest residential dwellings at Zolowee and so significant impacts on human receptors are not expected. The ore beneficiation process within the concentration plant is primarily a wet process and as such will not represent a significant source of dust. Misting and sprinkler systems are fitted to conveyors and transfer points. Tailings are pumped to the tailings management facility for emplacement as slurry, and will not generate dust emissions. The concentrate is sent via conveyor to a longitudinal stockpile, which will be reclaimed using a boom bucket reclaimer and sent to a surge bin for direct train loading, thereby minimising fugitive dust arisings. Front end loaders would not be needed for the loading process unless there are problems with the main train loading system. Overall, there are no known residential receptors within the areas that could be affected by mining operations and associated activities. Under normal climatic conditions, any dispersion of particulate matter emissions will be limited by the very low wind speeds recorded in this area, and it is expected that any impacts would be restricted to within 200 metres of the mine footprint and processing areas. The application of standard best practice measures for the mining, minerals extraction and ore processing activities within a formal EMP would be capable of controlling emissions to a level where effects on sensitive human receptors would not be significant. 29

31 5.1.3 Prediction of Effects on Crops and Sensitive Flora Any occasional dust plumes extending beyond the boundary of the mine area will be dependent upon local weather conditions, but given the very low wind speeds it is possible that vegetation may be affected by the deposition of dust at rates that exceed the baseline conditions within about 200 metres of the mine footprint boundary during the dry season. Again, the greatest potential for dust emissions will arise from the passage of trucks on haul roads. Any deposited material would be composed of soil like particles and any material that settled on vegetation would be readily displaced during periods of precipitation. Emissions during the wet season would be greatly reduced as dust emissions from unpaved roads would be substantially suppressed. The impact of dust deposition on sensitive terrestrial ecology and freshwater biodiversity is considered as necessary within relevant reports elsewhere within the ESIA, and are therefore not considered in this assessment. 5.2 Emissions of Combustion Pollutants from Power Plant Stacks This section reports the results of the dispersion modelling of emissions from the concentrator power plant, at the preferred stack height of 32 m, with 2% maximum sulphur fuel. Where it possible to do so, baseline pollutant concentrations relevant to the receptor in question have been taken from the project specific monitoring data presented in Section 4 of this report. Short term background concentrations (averaging period within a day) have been assumed to be double the long term concentration. Consideration of Stack Height The modelling study included an evaluation of the height of the power plant stacks, using the ADMS dispersion model. The selection of an appropriate stack height requires a number of factors to be taken into account, the most important of which is the need to balance a stack height sufficient to achieve adequate dispersion of pollutants against other constraints such as cost and visual impact. Emissions from the main stack have been modelled at stack heights between 20 m and 90 m. A graph, showing the process contribution (PC) to annual mean NO X concentrations is presented in Plot 5.1, below. The purpose of the graph is to evaluate the optimum stack height in terms of the dispersion of pollutants which would occur, against other constraints of further increases in release height. Analysis of the annual mean curve shows that the benefit of incremental increases in stack height up to 30 m is most pronounced. At heights above m, the curve flattens and the air quality benefit of increasing stack height further is reduced. It is therefore considered that 32 m represents a height at which site-specific constraints and the diminishing benefits of further increases in release height begin to outweigh the benefits to air quality. 30

32 Plot 5.1: Predicted Process Contribution to Maximum Annual Mean NO X Concentrations at Stack Heights between 20 m and 90 m Annual Mean NO2 PC (µg/m 3 ) Stack (m) Nitrogen Dioxide (NO 2 ) The predicted change in annual mean NO 2 concentrations that would occur during the operation of the power plant, at the selected sensitive receptors, is presented in Table 5.1. The predicted impact on short term NO 2 concentrations is presented in Table 5.2. Pollutant contour plots, showing the spatial distribution of impacts, are presented in Figure 5.1 and Figure 5.2. TABLE 5.1: PREDICTED IMPACT ON ANNUAL MEAN NO 2 CONCENTRATIONS Receptor Description Background Annual Mean NO 2 Concentration (µg/m 3 ) Process Contribution PEC PEC % AQS R1 Zolowee N R2 Zolowee S R3 Makinto R4 Geh R5 Gbapa R6 Vayanpa R7 Kpadapa Air Quality Standard (IFC) 40 31

33 The results in Table 5.1 show that the operation of the power plant would not cause a risk of exceeding the annual mean NO 2 air quality standard at the selected sensitive receptors where relevant human exposure would occur. Due to its proximity to the concentrator site, the most affected residential location would be Zolowee, where the predicted environmental concentration (PEC) would remain below 35% of the annual mean standard. As for the annual mean, the largest impact on short-term NO 2 concentrations would be seen in Zolowee. At this location and the other sensitive receptors considered within the assessment, there would not be a predicted exceedance of the 1-hour IFC standard. TABLE 5.2: PREDICTED IMPACT ON MAXIMUM 1 HOUR MEAN NO 2 CONCENTRATIONS Receptor Description Background 1 hour Mean NO 2 Concentration (µg/m 3 ) Process Contribution PEC PEC % AQS R1 Zolowee N R2 Zolowee S R3 Makinto R4 Geh R5 Gbapa R6 Vayanpa R7 Kpadapa Air Quality Standard (IFC) 200 Sulphur Dioxide (SO 2 ) (2% maximum S content in the fuel) The predicted change in 24-hour mean SO 2 concentrations that would occur during the operation of the power plant, at the selected sensitive receptors with 2% sulphur fuel, is presented in Table 5.3. The predicted impact on 10-minute SO 2 concentrations is presented in Table 5.4. Pollutant contour plots, showing the spatial distribution of impacts, are presented in Figure 5.3 and Figure 5.4. The results in Tables 5.3 and 5.4 show that, if 2% sulphur fuel is used for power generation, there would be a low risk of exceeding the short-term air quality standards for SO 2 at sensitive receptor locations. The predicted maximum 24-hour and 1-hour concentrations are around 35% and 78% of the air quality standards respectively. 32

34 TABLE 5.3: PREDICTED IMPACT ON 24 HOUR MEAN SO 2 CONCENTRATIONS Receptor Description Background 24 Hour Mean SO 2 Concentration (µg/m 3 ) Process Contribution PEC PEC % AQS R1 Zolowee N R2 Zolowee S R3 Makinto R4 Geh R5 Gbapa R6 Vayanpa R7 Kpadapa Air Quality Standard (IFC) 125 TABLE 5.4: PREDICTED IMPACT ON MAXIMUM 10 MINUTE MEAN SO 2 CONCENTRATIONS Receptor Description Background 10 Minute Mean SO 2 Concentration (µg/m 3 ) Process Contribution PEC PEC % AQS R1 Zolowee N R2 Zolowee S R3 Makinto R4 Geh R5 Gbapa R6 Vayanpa R7 Kpadapa Air Quality Standard (IFC)

35 Particulate Matter (PM 10 and PM 2.5 ) The predicted change in annual mean PM 10 concentrations that would occur during the operation of the power plant, at the selected sensitive receptors, is presented in Table 5.5. The predicted impact on short term PM 10 concentrations is presented in Table 5.6. The predicted change in annual mean PM 2.5 concentrations that would occur during the operation of the power plant, at the selected sensitive receptors, is presented in Table 5.7. The predicted impact on short term PM 2.5 concentrations is presented in Table 5.8. Pollutant contour plots, showing the spatial distribution of predicted impacts on PM 10 concentrations, are presented in Figure 5.5 and Figure 5.6. As the total particulate matter emission is assumed to be present as both PM 10 and PM 2.5, these contour plots apply to the predicted change in PM 2.5 also. The results tables show that the impact on local concentrations of particulate matter from the operation of the power plant would be very low in comparison to existing baseline concentrations. Although the PM 10 air quality standards are predicted to be exceeded or at risk of exceedance, this is overwhelmingly due to existing concentrations rather than the predicted impact of the operational power plant. Once the upgrade of the Sanniquellie to Yekepa highway is completed, it is likely that overall particulate matter concentrations will decrease significantly over those measured during the baseline survey. TABLE 5.5: PREDICTED IMPACT ON ANNUAL MEAN PM 10 CONCENTRATIONS Receptor Description Background 1 Hour Mean PM 10 Concentration (µg/m 3 ) Process Contribution PEC PEC % AQS R1 Zolowee N R2 Zolowee S R3 Makinto R4 Geh R5 Gbapa R6 Vayanpa R7 Kpadapa Air Quality Standard (IFC) 70 34

36 TABLE 5.6: PREDICTED IMPACT ON 99 TH PERCENTILE 24 HOUR MEAN PM 10 CONCENTRATIONS Receptor Description Background 24 Hour Mean PM 10 Concentration (µg/m 3 ) Process Contribution PEC PEC % AQS R1 Zolowee N R2 Zolowee S R3 Makinto R4 Geh R5 Gbapa R6 Vayanpa R7 Kpadapa Air Quality Standard (IFC) 150 TABLE 5.7: PREDICTED IMPACT ON ANNUAL MEAN PM 2.5 CONCENTRATIONS Receptor Description Background 1 Hour Mean PM 2.5 Concentration (µg/m 3 ) Process Contribution PEC PEC % AQS R1 Zolowee N R2 Zolowee S R3 Makinto R4 Geh R5 Gbapa R6 Vayanpa R7 Kpadapa Air Quality Standard (IFC) 35 35

37 TABLE 5.8: PREDICTED IMPACT ON 99 TH PERCENTILE 24 HOUR MEAN PM 2.5 CONCENTRATIONS Receptor Description Background 24 Hour Mean PM 2.5 Concentration (µg/m 3 ) Process Contribution PEC PEC % AQS R1 Zolowee N R2 Zolowee S R3 Makinto R4 Geh R5 Gbapa R6 Vayanpa R7 Kpadapa Air Quality Standard (IFC) 75 Carbon Monoxide (CO) The predicted change in 8-hour mean CO concentrations that would occur during the operation of the power plant, at the selected sensitive receptors, is presented in Table 5.9. Although no baseline data is available, due to the lack of large combustion sources in the area it is very unlikely that the air quality standard would be exceeded. TABLE 5.9: PREDICTED IMPACT ON 8 HOUR MEAN CO CONCENTRATIONS Receptor Description Background 8 Hour Mean CO Concentration (µg/m 3 ) Process Contribution PEC PEC % AQS R1 Zolowee N <0.1 R2 Zolowee S <0.1 R3 Makinto <0.1 R4 Geh <0.1 R5 Gbapa <0.1 R6 Vayanpa <0.1 R7 Kpadapa <0.1 Air Quality Standard 10,000 Maximum Impacts within the Modelled Domain The maximum Process Contribution (PC) and Predicted Environmental Concentration (PEC) within the modelled domain, including locations where there is no relevant human exposure, 36

38 are summarised in Table 5.10 for each pollutant and averaging period. The distribution of impacts is also illustrated in Figure 5.1 to Figure 5.6 inclusive. TABLE 5.10: 32M STACK HEIGHT WITH 2% SULPHUR FUEL - MAXIMUM PROCESS CONTRIBUTION AND PREDICTED ENVIRONMENTAL CONCENTRATION, ALL MODELLED POLLUTANTS Pollutant Averaging Period AQS PC PC PEC PEC (µg/m 3 ) (µg/m 3 ) (% EAL) (µg/m 3 ) (% EAL) Annual Mean NO 2 Maximum 1-hour mean Annual Mean PM th %ile of 24-hour means Annual Mean PM 2.5 SO 2 CO 99 th %ile of 24-hour means Maximum 24-hour mean Maximum 10-min mean Max daily 8-hour running mean , n/a n/a It can be seen that the operation of the power plant would cause an exceedance of the 10- minute SO 2 air quality standard with a 32 m stack. The dispersion of pollutants is greatly influenced by the local terrain, and in this case the predicted impact is centred on an area near to the peak of Mount Tokadeh within the mine site, which can be seen on the pollutant contour plot. As this area is within the exclusion zone members of the public would not be present at this location, it is therefore more relevant to consider impacts where human exposure may occur. The exclusion zone would also prevent farmers from entering the immediate area around mining activities. The PC to PM 10 concentrations would not directly cause an exceedance of the air quality standards at any location within the modelled domain, although high baseline concentrations at some locations would mean that the air quality standards would be exceeded both with and without the power plant in operation. There would not be a predicted exceedance of the annual mean and 1-hour air quality standards for NO 2, the 24-hour standard for SO 2 or the standard for CO at any location within the modelled domain. IFC General EHS guidelines state that to allow for future sustainable development within the same airshed, emissions from a single project should not normally contribute more than 25% of a relevant air quality standard. In this case, however, it is considered that there would not be a significant chance of extensive further industrial development in the immediate vicinity of Tokadeh over the life of the project. Therefore, the proposed development does not have the potential to restrict further sustainable development in the area. Consideration of Power Plant Operation with 4% Sulphur Fuel 37

39 In addition to a consideration of SO 2 impacts from the power plant operating with fuel containing the IFC guideline limit of 2% sulphur, the dispersion modelling assessment has also evaluated the impact of the use of fuel containing 4% sulphur on stack height. Table 5.11 presents a comparison of maximum predicted 24-hour PC values for a range of stack heights at the most affected sensitive receptor (R1, N Zolowee). The results are reported both as the PC in µg/m 3 and as a percentage of the IFC guideline air quality standard. A comparison of predicted impacts on maximum 10-minute concentrations is presented in Table TABLE 5.11: COMPARISON OF PREDICTED IMPACTS ON MAXIMUM 24 HOUR MEAN SO 2 CONCENTRATIONS AT NORTH ZOLOWEE Stack Height PC with 2% S 24-hour Mean SO 2 Concentration (µg/m 3 ) % AQS with 2% S PC with 4% S % AQS with 4% S 20 m m m m m m m m m m m m m m m m m m m m m m

40 TABLE 5.12: COMPARISON OF PREDICTED IMPACTS ON MAXIMUM 10 MINUTE MEAN SO 2 CONCENTRATIONS AT NORTH ZOLOWEE Stack Height PC with 2% S 10 Minute Mean SO 2 Concentration (µg/m 3 ) % AQS with 2% S PC with 4% S % AQS with 4% S 20 m m m m m m m m m m m m m m m m m m m m m m Although Table 5.11 shows that the 24-hour standard is unlikely to be exceeded with either fuel at Zolowee, the PC to 10-minute mean concentrations would be such that the predicted PC would exceed the air quality standard with 4% sulphur fuel at stack heights up to approximately 65 m. The PC would be more than 90% of the standard with 4% sulphur fuel at stack heights up to approximately 80 m. For this reason, it is recommended that the sulphur content of HFO fuel specified for use in the concentrator power plant is maintained below 2%, as per IFC guidelines. If supplies of HFO with sulphur contents in this range cannot be guaranteed by the supplier, then there is a risk that the short term air quality standard for SO 2 could be exceeded in Zolowee with a stack height of less than 80 m above ground level. 39

41 5.3 Railway Corridor Dust Emissions from Trains There is potential for dust emissions to arise from payload on open cars, as the ore would be loaded to above the level of the top of the wagon sides and the ore is transported uncovered. Currently, Phase 1 DSO ore trains consist of up to 80 cars hauled by a single locomotive, and there are up to two trains per day in each direction. For Phase 2, the completion of the extended loops at Buchanan will allow trains to be increased to 140 cars hauled by two locomotives, with up to three trains per day in each direction. The maximum speed of the train is 60km/hr when loaded. The loading and transport of DSO was observed during a visit to the Tokadeh mine site and railway corridor during the dry season in March As the beneficiated ore from the concentrator would be of a similar nature in terms of size and moisture content, the likely impacts associated with the transport of ore produced during Phase 2 would be comparable with those observed with DSO. Due to the inherent moisture content and density of the ore, it was seen that there was a low potential for dust emissions to occur from the cars during the loading process, with no evidence of significant emissions of fine material to atmosphere. Once underway, the load was observed to be stable and no evidence of significant dust emissions from the cars was observed while the train was travelling at maximum speed. Similarly, there was no evidence that the deposition of ore dust was occurring on vegetation or property surfaces at locations along the railway corridor. Given the low frequency of railway movements and the observed stability of the ore during loading and transport, significant impacts due to fugitive emissions of ore from open railway cars are therefore not considered likely. Any deposition of ore material would be infrequent, low in magnitude and restricted to locations in very close proximity to the route of the railway. Ore dust would be readily displaced during periods of precipitation. Fugitive Emissions of Dust from the Service Road Project related traffic on the service road is likely to reduce upon completion of railway rehabilitation. There would be an increase in movements between Yekepa and the mine sites associated with the transportation of mine workers and goods, which could increase dust emissions. ArcelorMittal Liberia Limited has pledged funding to the Liberian government to upgrade the Yekepa Ganta highway to a national blacktop standard. This would significantly reduce dust emissions arising from re-suspension along this entire route, and would have a beneficial impact to communities in Yekepa, Gbapa, Zolowee and Sanniquellie. South of km 255, there would be continued use of the service road throughout the operational period of the mine. All heavy materials etc would be transported along the railway, but workers and tools etc would be transported by road. During dry periods of weather, vehicles passing over sections of the unpaved roads cause significant re-suspension of dust, and without mitigation there could be a significant impact on human populations living close to the route. At locations further from the roadside, in areas where thick vegetation screening is likely to be in place between the source and receptors, the majority of re-suspended or windblown dust from roads is deposited within 50 m to 100 m of the source of the emission. Smaller scale dust effects of minor significance could occur at distances of up to 200 metres. 40

42 Greenhill Quarry At Greenhill Quarry, the principal sources of dust are: Vehicle movements on the site access; Drilling and blasting of within the quarry; Excavation of blasted material and transport to the processing area; Crushing, screening and stockpiling within the processing area; and Loading of rock onto rail wagons for onward transport. The closest population settlements to the quarry are situated adjacent to the access road. No properties are situated within 200 m of the quarry or processing and loading areas, so there is limited potential for dust impacts to occur from these sources. Blasting operations at the site are infrequent and occur on average about once per month. There are, therefore, a small number of sensitive receptors within the areas that could be affected by on site operations. The greatest potential for fugitive dust emissions is likely to arise from the passage of trucks and other road vehicles along the access road outside the site. This is highly unlikely to represent a problem during the wet months, when there is consistently more than 1mm of rainfall per day. In the absence of suitable mitigation measures, however, there is the potential for significant impacts on human populations close to the road to occur during periods of dry weather. 5.4 Port (Landside) Dust Effects Any effects associated with the construction of Phase 2 landside rehabilitation works would be occasional and temporary in nature. The majority of construction activities would take place towards the centre of the port, at a distance of around 500 m from the nearest human settlements in Buchanan, and are highly unlikely to give rise to significant impacts. The installation of the new rail loop and railcar dumper building would however take place within 100m of the boundary of the site, and suitable mitigation would therefore need to be employed in these areas to prevent significant impacts occurring. Even with the application of mitigation, occasional impacts cannot be discounted. In terms of port operations, the stocking area and stacker reclaimer would be situated towards the centre of the port, around 500 m from the nearest residential receptors. Significant impacts on dust deposition rates and particulate matter concentrations would therefore be unlikely to occur at such a distance from the source. Ship loading operations would take place at the southern end of the port development, more than 1 km from the nearest human receptors. The proposed location of the railcar dumper building is within 50 m of the western boundary of the port, and within around 100 m of the nearest residential properties. The dumper itself would be partially enclosed within a steel clad building. During tippler operations, there would be a low overall potential for significant emissions of dust and particulate matter to occur due to the inherent moisture content of the ore and the enclosure of the tipping mechanism. The laydown yard would be between metres distant from the closest settlements. Based on the distance criteria set out in Section 1.1.2, significant effects could occur at these closest properties unless appropriate mitigation is applied. 5.5 Emissions of Greenhouse Gases The Scheme (along with any other development) inevitably increases emissions of carbon dioxide. However, carbon dioxide emissions in Liberia represent a very small proportion of those within the sub-saharan region, and the Scheme during Phase 2 is very unlikely to change this overall position. The use of the railway for the transportation of ore from the mining areas to Buchanan Port would take place using the rehabilitated railway line, which represents the lowest carbon footprint for movement of the ore. 41

43 6 MITIGATION 6.1 Principles The general principles of mitigation and environmental design within the Scheme would be consistent within the hierarchy of measures set out within this ESIA: Avoidance such as the use of techniques or design measures which prevent impacts from occurring; Minimisation implementing measures or design features which reduce undesirable impacts; Rectification rehabilitation or restoration of affected areas; and Compensating for the impact by replacing or providing substitute resources. Compensation should be seen as a last resort. A number of standard measures can be implemented which are common practice on mining and minerals extraction projects around the world. The emphasis in the mitigation of impacts on air quality would be placed on the avoidance or minimisation of effects wherever possible within the scheme design. Mitigation would be focused on the following areas: The avoidance of dust generation through the use of appropriate techniques, such as the use of bulk rail transport over road haulage; Avoidance of exposure by spatial design of site and worker settlements; Minimisation of dust generation at source through design; Minimisation of dust generation at source through the effective management of site operations; and Active management of dust generation, for example through water suppression, screening, etc. to minimise potential for adverse impacts. 6.2 Specific Measures Detailed measures for the control of fugitive emissions from the extraction and processing areas, transport corridors and port operations should be set out within the Scheme Environmental Management Plan (EMP). A summary of the measures is outlined within this section Mine Area The restriction of movements of topsoil and overburden to avoid unnecessary handling of materials; Considerate siting of stockpiles to be as far away from site boundary as feasible; Implement speed limits on haul routes; Water spray haul routes, handling and loading areas as necessary during periods of drier weather; Minimisation of drop heights from excavators and payloaders; Maintain vegetative screening between stockpile/loading areas and human settlements. These should be designed to provide mixed, dense vegetation, and should be located close to the dust-emitting sources; and Emissions from all mobile plant to comply with US EPA Tier 3 emissions limits. 42

44 6.2.2 Ore Processing and Concentrator Area Emissions for stationary combustion plant to comply with IFC Environmental, Health and Safety Guidelines and emission limits; Water suppression and extract fans/bag filters to be employed as necessary on the ore crusher and concentrator; Use of water spray cannons on the ore stockpile area; and Use of a boom bucket reclaimer and surge bin on the rail load out facility, in preference to loading shovels Railway Corridor Port (landside) Restrict locomotive idling at loading and unloading depots, and at any other locations; Maintain strict controls on overloading of rail cars; All main line locomotives to comply with US EPA Tier 2 emissions limits; and Restrict vehicle speeds on sections of unpaved roads which pass through populated areas. Restrict locomotive idling within the port area; Use of water-sprays to ensure that any areas of hard standing (including the laydown yard) and unpaved areas are maintained in a damp condition when in use; Sitting of the stocking area and stacker reclaimer towards the centre of the port, away from the site boundary; Enclosure of the rail tippler to minimise fugitive emissions during unloading; and Use of a stacker reclaimer and conveyor system for ship loading 43

45 7 RESIDUAL EFFECTS 7.1 Mine and Processing Areas There are no residential receptors within the areas that could be significantly affected by dust emissions from mining and ore processing activities. Under normal climatic conditions, any dispersion of particulate matter emissions will be limited by the low wind speeds recorded in the area. The application of standard best practice measures for the mining and minerals extraction sector within a formal EMP would be capable of controlling emissions to a level where effects on sensitive human receptors would not be significant. 7.2 Emissions of Combustion Pollutants from Power Plant Stacks A stack height of 32 m would be sufficient to ensure there would not be a breach of ambient air quality standards at sensitive receptor locations, if the plant is operated within emissions limits specified by IFC guidance. If the sulphur content of the HFO fuel cannot be guaranteed by the supplier to be below the IFC standard of 2%, then there is risk that the short term air quality standard for SO 2 could be exceeded at Zolowee with 4% fuel, for stack heights up to 80 m. 7.3 Railway Corridor Given the low frequency of railway movements, significant effects on human populations due to operational rail movements are not considered likely. The mitigation measures proposed would give further confidence that emissions from railway operations can be controlled to a level where the effects are of negligible significance. The paving of sections of road where they pass through populated areas, and the upgrade of the Yekepa Ganta highway to a national blacktop standard would significantly reduce emissions arising from dust re-suspension along the entire railway corridor route. Following the completion of the upgrade programme, the effect of additional road traffic on sensitive human receptors would not give rise to a significant effect at air quality sensitive receptors. In areas where permanent road sealing is not planned, project related traffic could cause significant re-suspension of dust during dry weather. These emissions would contribute to possible significant effects on human populations living within 50 m to 100 m of the route. Limiting the speed of project related traffic travelling along unsealed roads would reduce the magnitude of the impact. 7.4 Greenhill Quarry There are no residential properties within 200 m of the quarry or processing and loading areas, so there is limited potential for dust impacts to occur from these sources. Properties close to the unsealed access road may however be affected by re-suspended road dust during dry weather. Limiting the speed of project related traffic would reduce the magnitude of the impact. 7.5 Port (landside) Although there are settlements located within 200 m to 300 m of the laydown area, and within 100 m of the railcar dumper, the application of appropriate mitigation would be capable of reducing effects at sensitive receptors off site to acceptable levels. Occasional increases in dust deposition rates at small numbers of receptors may occur when activities are carried out during times of very dry and windy weather, however during these times background concentrations of particulate matter would also become elevated and the overall effect would be of slight adverse significance. 7.6 Greenhouse Gas Emissions Liberia represents a very small proportion of the total CO 2 emissions within the sub-saharan region, and the Phase 2 mining, transport and shipping operation would be very unlikely to change this position. The proposed mitigation would minimise CO 2 emissions to the lowest level possible. 44

46 8 CONCLUSIONS This report has assessed the impact of the proposed Scheme on air quality. The Scheme would have the potential to change ambient concentrations of airborne particulate matter and rates of dust deposition in the area around the mine extraction sites, crusher, concentrator, rail loading facility, the railway corridor and Buchanan Port. There are no residential receptors within the areas that could be significantly affected by dust emissions from mining and ore processing activities. Under normal climatic conditions, any dispersion of particulate matter emissions will be limited by the low wind speeds, indicated by the records in the area. The application of standard best practice measures for the mining and minerals extraction sector within a formal EMP would be capable of controlling emissions to a level where effects on sensitive human receptors would not be significant. Dispersion modelling of emissions from the Tokadeh concentrator power plant has demonstrated that a stack height of 32 m would be sufficient to ensure there would not be a breach of ambient air quality standards at sensitive receptor locations, if the plant is operated within emissions limits specified by IFC guidance. Modelling of the plant operating on 4% sulphur content fuel, however, predicts that the PC would be more than 100% air quality standard at stack heights up to approximately 65 m, and more than 90% of the standard with at stack heights up to approximately 80 m. For this reason, it is recommended that the sulphur content of HFO fuel specified for use in the concentrator power plant is maintained below 2%, as per IFC guidelines. If supplies of HFO with sulphur contents in this range cannot be guaranteed by the supplier, then there is a risk that the short term air quality standard for SO 2 could be exceeded in Zolowee with a stack height of less than 80 m above ground level. Given the low frequency of railway movements, significant effects on human populations due to operational rail movements are not considered likely. The mitigation measures proposed would give further confidence that emissions from railway operations can be controlled to a level where the effects are of negligible significance. Measures to seal sections of road where they pass through populated areas, and the upgrade of the Yekepa to Ganta highway to a national blacktop standard would significantly reduce emissions arising from dust re-suspension from local roads. Following the completion of the upgrade programme, the effect of additional road traffic on sensitive human receptors would not give rise to a significant effect at air quality sensitive receptors. In areas where permanent road sealing is not planned, project related traffic could cause significant re-suspension of dust during dry weather. Without mitigation, these emissions could contribute to significant effects on human populations living close to the route. Further from the roadside, thick vegetation screening reduces such impacts so that the majority of the dust is deposited within 50 m to 100 m of the source. Limiting the speed of project related traffic travelling along unsealed roads would reduce the magnitude of impacts to a level that would give an effect of slight adverse significance. At Greenhill Quarry, there are no residential properties within 200 m of the quarry or processing and loading areas, so there is limited potential for dust impacts to occur from these sources. Properties close to the unsealed access road may however be affected by resuspended road dust during dry weather, and limiting the speed of project related traffic would reduce the magnitude of this impact to a level that would give an effect of slight adverse significance. Although there are settlements located within 200 m to 300 m of the laydown area, and within 100 m of the railcar dumper, the application of appropriate mitigation would be capable of reducing effects at sensitive receptors off site to acceptable levels. Occasional increases in dust deposition rates at small numbers of receptors may occur at times when activities are carried out during times of very dry and windy weather, however during these times background concentrations of particulate matter would also become elevated and the overall effect would be of slight adverse significance. 45

47 Liberia represents a very small proportion of the total CO 2 emissions within the sub-saharan region, and the Phase 2 Scheme would be very unlikely to change this position. The proposed mitigation would minimise CO 2 emissions to the lowest level possible. Overall, the impact of the Scheme can be mitigated such that it would not represent a significant effect. 46

48 9 REFERENCES CERC, 2007, ADMS 4 Validation Papers Department of the Environment (1995) The Environmental Effects of Dust from Surface mineral Workings, Volume 2, Arup Environmental and Ove Arup and Partners European Commission (2001), Integrated Pollution Prevention and Control, Best Available Techniques Reference Document on the Production of Iron and Steel. Folkins, I et al. Tropical ozone as an indicator of deep convection, Journal of Geophysical Research, Vol. 107, No. D13, 4184, /2001JD001178, 2002 GoL (2003) An Act Adopting the Environmental Protection and Management Law of the Republic of Liberia. Published by authority, April 30, 2003, Monrovia: Ministry of Foreign Affairs IFC (2006) International Finance Corporation s Performance Standards on Social and Environmental Sustainability, Washington D.C.: IFC/World Bank Group IFC (2007a) International Finance Corporation s Guidance Notes: Performance Standards on Social & Environmental Sustainability, Washington DC: IFC/World Bank Group IFC (2007b) Environmental, Health, and Safety General Guidelines, Washington DC: IFC/World Bank Group IFC (2007c) Environmental, Health, and Safety Guidelines for Thermal Power Plants, Washington DC: IFC/World Bank Group ODPM (2005). Minerals Policy Statement 2: Controlling and Mitigating the Environmental Effect of Minerals Extraction in England Annex 1: Dust. London: ODPM Publications SANS 1929 (2004). South African National Standard: Ambient Air Quality Limits for Common Pollutants. Pretoria, Guateng, SA: South African Bureau of Standards UN (1992a) United Nations Framework Convention on Climate Change of 16 th February 2005 (entered into force). Bonn: UNFCCC Secretariat UN (1992b) United Nations Framework Convention on Climate Change of 16 th February 2005 (entered into force). Bonn: UNFCCC Secretariat UNEP (1985) The Vienna Convention for the Protection of the Ozone Layer. Ozone. Nairobi Secretariat / UNEP USEPA (1998). Compilation of Air Pollution Emissions Factors: EPA-42. Fifth Edition, Volume I, Chapter 11. WHO (2000). Air Quality Guidelines for Europe. 2 nd ed. (European Series: No. 91). Copenhagen: World Health Organisation Regional Office for Europe. WHO (2005). Air Quality Guidelines Global Update Report on a Working Group Meeting, Bonn, Germany, October

49 APPENDIX 1: FIGURES 48

50 APPENDIX 2: BUCHANAN POWER PLANT DISPERSION MODELLING ASSESSMENT 49

51 50

52 51

53 52

54 53

55 54

56 55

57 Nimba Western Area Iron Ore Concentrator Mining Project, Liberia Environmental and Social Impact Assessment Appendix 2 to Volume 3, Part 5: Buchanan Power Plant Dispersion Modelling Assessment March 2013 Prepared for: ArcelorMittal Liberia Limited

58 REVISION SCHEDULE Rev Date Details Prepared by Reviewed by Approved by 1 13 September 2012 Draft Report Danny Duce Principal Air Quality Consultant Chuansen Ren Senior Air Quality Specialist Martin Edge Regional Director Europe & Africa Gareth Hearn Project Technical Director 2 11 October 2012 Final Draft Danny Duce Principal Air Quality Consultant Gareth Hearn Project Director Tanya Romanenko Martin Edge Regional Director Europe & Africa Project Manager 3 March 2013 Final Danny Duce Principal Air Quality Consultant Tanya Romanenko Project Manager Martin Edge Regional Director Europe & Africa URS Scott House Alençon Link Basingstoke Hampshire RG21 7PP Tel +44 (0) Fax +44 (0) VOLUME 3, PART 5, APPENDIX 2: BUCHANAN DISPERSION MODEL 1

59 Limitations URS has prepared this Report for the sole use of ArcelorMittal Liberia Limited ("Client") in accordance with the Agreement under which our services were performed and may not be relied upon by any other party without the prior and express written agreement of URS. No other warranty, expressed or implied is made as to the professional advice included in this Report or any other services provided by URS. The methodology adopted and the sources of information used by URS in providing its services are outlined in this Report. The work described in this Report was undertaken variously between 2010 and 2011 for Phase 1 and 2011 to June 2012 for Phase 2 and is based on the conditions encountered and the information made available by the Client during the said period of time. The scope of this Report and the services are accordingly factually limited by these circumstances. The layouts shown in supporting Volumes 3 to 5 inclusive have been superseded. The final layouts (December 2012) are provided and assessed in Volume 1 only. Where assessments of works or costs identified in this Report are made, such assessments are based upon the information available at the time and where appropriate are subject to further investigations or information which may become available. URS disclaim any undertaking or obligation to advise any person of any change in any matter affecting the Report, which may come or be brought to URS attention after the date of the Report. Certain statements made in the Report that are not historical facts may constitute estimates, projections or other forwardlooking statements and even though they are based on reasonable assumptions as of the date of the Report, such forward-looking statements by their nature involve risks and uncertainties that could cause actual results to differ materially from the results predicted. URS specifically does not guarantee or warrant any estimate or projections contained in this Report. Where field investigations are carried out, these have been restricted to a level of detail required to meet the stated objectives of the services. The results of any measurements taken may vary spatially or with time and further confirmatory measurements should be made after any significant delay in issuing this Report. Copyright This Report is the copyright of the Client. Any unauthorised reproduction or usage by any person other than the Client is strictly prohibited. VOLUME 3, PART 5, APPENDIX 2: BUCHANAN DISPERSION MODEL 2

60 TABLE OF CONTENTS 1 INTRODUCTION Overview Assessment Scope Air Quality Criteria and Standards METHODOLOGY Baseline Determination Air Quality Sensitive Receptors Survey of Baseline Nitrogen Dioxide and Sulphur Dioxide Meteorological Conditions and Topography Prediction and Evaluation of Air Quality Effects Dispersion Modelling of Port Power Plant Emissions. 6 3 BASELINE CONDITIONS Receptor Locations Measured Dust and Particulate Matter Concentrations Measured Baseline Nitrogen Dioxide and Sulphur Dioxide DISPERSION MODELLING RESULTS CONCLUSIONS REFERENCES APPENDIX: FIGURES VOLUME 3, PART 5, APPENDIX 2: BUCHANAN DISPERSION MODEL 3

61 1 INTRODUCTION 1.1 Overview This addendum report considers the potential effects of emissions to air from the operation of the port power plant at Buchanan, on surrounding air quality sensitive receptors. The assessment of the power plant does not form part of the main Phase 2 ESIA. An assessment of the air quality effects of the other elements of the Phase 2 Scheme, arising from changes in air quality associated with dust generation and emissions of gaseous pollutants and fine particulate matter, can be found within the main body of the ESIA. 1.2 Assessment Scope The main air quality impact during the operation of the proposed port power plant would be emissions to air from the combustion of fuel within the power plant engines. The primary fuel for the plant will be Heavy Fuel Oil (HFO). There are currently no national limits for emissions from power plants in Liberia. Therefore emission guidelines for new thermal power plants burning fossil fuels, as detailed in the IFC guidelines for Thermal Power Plants, have been employed in the evaluation of the facility. The air quality impacts of the proposed power plant have been assessed using the plume dispersion model ADMS, which is able to calculate maximum ground level concentrations at sensitive receptors close to the facility site. The dispersion model has been used to verify the proposed stack height as appropriate, and to demonstrate that the predicted impacts of the operation of the plant at air quality sensitive receptors are acceptable in the context of WHO air quality standards. 1.3 Air Quality Criteria and Standards The air quality standards, against which the impact of power plant operation has been assessed, are shown in Table 1.1, below. In proposing air quality standards for the project, it is important to draw a distinction between the guidelines recommended by WHO, and the standards that have been brought into legislation by e.g. the European Union. The former are purely based on the scientific and medical evidence of the effects of an individual pollutant. The latter take into account the extent to which the standards are expected to be achieved by a certain date including economic efficiency, practicability, existing air quality conditions and technical feasibility and timescale. TABLE 1.1: AIR QUALITY STANDARDS ADOPTED FOR THE ESIA Pollutant Averaging Period Standard Standard Derived From Sources PM hours 150 µg/m 3 IFC (adopted from Mining operations, Annual (99 th percentile) 70 µg/m 3 WHO Guidelines, Interim Target 1) vehicle exhausts, railway locomotives, mean power generation PM hours 75 µg/m 3 (99 th percentile) Annual mean 35 µg/m 3 Nitrogen 1 hour 200 µg/m 3 IFC (adopted from Vehicle exhausts, dioxide (NO 2 ) Annual mean 40 µg/m 3 WHO Guidelines) railway locomotives, power generation Sulphur dioxide (SO 2 ) 10 min mean 24 hours 500 µg/m µg/m 3 IFC (adopted from WHO Guidelines) IFC (adopted from WHO Guidelines, Interim Target 1) Railway locomotives, heavy fuel oil-fired power generation Carbon monoxide (CO) 8 hours 10 mg/m 3 WHO Guideline Vehicle exhausts, railway locomotives, power generation PM 10 and PM 2.5 is fine particulate matter with an aerodynamic diameter of less than 10 and 2.5 micrometres respectively VOLUME 3, PART 5, APPENDIX 2: BUCHANAN DISPERSION MODEL 4

62 2 METHODOLOGY 2.1 Baseline Determination Air Quality Sensitive Receptors Human Receptors An extensive survey of settlements and villages in the area around the port was undertaken in advance of Phase 1 of the project, during The identification of residential areas and individual properties potentially affected by emissions to air was aided by: Interpretation of aerial photography to identify dwellings within varying distance bands from sources of dust emissions; and Ground truthing of potential receptors through fieldwork undertaken by the ArcelorMittal Liberia and / or their Consultants during the baseline data collection for Phase 1 ESIA. to confirm the number, nature and level of use. Since the production of the Phase 1 ESIA, further visits to the port and surrounding area have been made by the URS Phase 2 ESIA project team. Recent information gained during the site visits has been combined with the existing Phase 1 information Survey of Baseline Nitrogen Dioxide and Sulphur Dioxide Emissions from the power plant to be constructed at the Port at Buchanan would have the potential to increase local concentrations of NO 2 and SO 2. In order to evaluate baseline concentrations of these two pollutants, a project specific passive diffusion tube monitoring survey has commenced in the area around the port site. The sampling is programmed to run for six months and was initiated during March The survey will therefore capture data in both the wet and dry seasons and, once complete, will be representative of annual mean baseline conditions. Passive diffusive samplers are simple devices which are widely used for the measurement of ambient NO 2 and SO 2 concentrations. Whilst they do not offer the same precision or accuracy as automatic chemiluminescent monitoring, their robust nature and ease of use make them a useful tool to evaluate long term pollutant concentrations. The diffusion tubes were placed at three locations in the area around Buchanan. Site B1 was selected to be representative of background conditions in rural and residential areas, while B2 and B3 were chosen to be representative of human exposure in locations around the port site. The sites selected for the survey are listed in Table 2.1. TABLE 2.1: DIFFUSION TUBE MONITORING LOCATIONS FOR NO 2 AND SO 2 Site ID Location Descriptor Grid Reference B1 AML office at Buchanan , B2 Moore Town Gate, Buchanan , B3 North of Buchanan Port Gate , Meteorological Conditions and Topography In order to inform the prediction of air quality it is necessary to determine the dispersion characteristics in the vicinity of emission sources. Hourly sequential meteorological data is required for use in dispersion modelling studies, and it is important that the data used are representative of conditions at the point of emission and the surrounding modelled domain. Data generated from the weather station installed at Buchanan has been used in this study. The station is located at Grid Ref: , VOLUME 3, PART 5, APPENDIX 2: BUCHANAN DISPERSION MODEL 5

63 2.2 Prediction and Evaluation of Air Quality Effects Dispersion Modelling of Port Power Plant Emissions Overview This section describes the dispersion model used to evaluate the impact of emissions to air on ambient pollutant concentrations in the vicinity of the port power plant site. Dispersion Model Selection The air quality impacts of the proposed power plant are best evaluated using a refined, near field (less than 50 km from the emission source) Gaussian Plume Dispersion Model, which is able to calculate maximum ground level concentrations at receptors close to the plant stacks. Gaussian models assume that pollutants do not decompose in the atmosphere, and therefore do not account for the long-range transport of atmospherically reactive pollutants. They are designed to produce results that are close to monitored values. The assessment of emissions from the power plant stacks has been undertaken using ADMS 4.2, supplied by Cambridge Environmental Research Consultants Limited (CERC). ADMS is a modern dispersion model that has an extensive published validation history (CERC, 2007). In particular, this version of ADMS has proven benefits for modelling meteorological conditions where there is a high incidence of calm conditions. Modelled Scenarios The dispersion modelling has been undertaken for a single operational scenario. The power plant would contain five engines, and the assessed scenario consists of all five engines running continuously without any emission variations. Due to the current uncertainty surrounding the guaranteed sulphur content of fuel available for delivery to Buchanan, the assessment has considered two scenarios for SO 2 : one scenario in which the plant would be operated according to IFC guidelines with maximum 2% sulphur fuel, and one in which the plant would be operated with 4% sulphur fuel (the current guaranteed level given by the fuel supplier). Dispersion Model Inputs The general model conditions used in the assessment are summarised in Table 2.2. Other more detailed data used to model the dispersion of emissions is considered below. TABLE 2.2: GENERAL ADMS MODEL CONDITIONS Variable Surface Roughness at source Receptors Receptor Location Source location Emissions Sources Meteorological data Terrain data Buildings that may cause building downwash effects Input 0.5 m Gridded and selected discrete receptors x,y co-ordinates determined by GIS, z = 1.5 m x,y co-ordinates determined by GIS Data provided by Caterpillar and AMEC project team 5 main exhaust stacks in 1 cluster Hourly sequential dataset from ArcelorMittal station at Buchanan (period 1/07/2011 to 30/06/2012) Flat terrain Turbine Hall VOLUME 3, PART 5, APPENDIX 2: BUCHANAN DISPERSION MODEL 6

64 Emissions Data The physical properties of the combustion emission sources, as represented within the model, are presented in Tables 2.3 and 2.4. This data has been taken from the scheme design drawings for the port development and emissions information provided by the AMEC project team and the engine manufacturers, Caterpillar. TABLE 2.3: MODELLED EMISSION SOURCES Release Point Grid Reference 9CM32_ , CM32_ , CM32_ , CM32_ , CM32_ , Note: The x and y coordinates listed are referenced to the WGS co-ordinate system, UTM zone 29N VOLUME 3, PART 5, APPENDIX 2: BUCHANAN DISPERSION MODEL 7

65 TABLE 2.4: PHYSICAL PROPERTIES, COMBUSTION PLANT STACKS (PER ENGINE) Parameter Unit Value Stack Height (5 flues) m 32 Effective internal stack diameter m 0.8 Flue temperature C 300 Volume flux m 3 /s 15.0 Stack gas velocity m/s 29.8 NO X mass emission rate g/s 14.0 PM 10 mass emission rate g/s 0.48 SO 2 mass emission rate (2% S) g/s 11.2 SO 2 mass emission rate (4% S) g/s 22.4 CO mass emission rate g/s 0.77 The modelled pollutant emission rates (in g s -1 ) are determined by the emissions limits set out within Table 6(A) of the IFC guidance for Thermal Power Plants (Emissions Guidelines for Reciprocating Engines of bore size less than 400 mm, Liquid Fuels, Plant >50 MWth to <300 MWth). These limits are: Particulate Matter: 50 mg/m 3 Sulphur Dioxide (SO 2 ): 1,170 mg/m 3 or <2% S in fuel Oxides of Nitrogen (NO 2 ): 1,460 mg/m 3 Carbon Monoxide (CO): 70 mg/m 3 Permitted pollutant mass emission rates from the power plant engines (in g/s) were calculated by multiplying the IFC emission limit concentrations by the volumetric flow rate at full load operation (at reference conditions of 15% O 2 ). For the 4% fuel sulphur scenario, the mass emission rate has been assumed to be 2 times the value for 2% sulphur fuel supplied by Caterpillar. Modelled Domain Ground-level concentrations of the modelled pollutants have been predicted at 17 discrete air quality sensitive receptors, as listed in Table 2.5. The receptors have been selected as locations where the nearest residential properties are found in each direction from the power plant site. The receptor locations have been selected to be representative of residential properties in the area around the site, and can be considered to be representative of other receptors in their vicinity. The height of the receptors has been set at 1.5 m. VOLUME 3, PART 5, APPENDIX 2: BUCHANAN DISPERSION MODEL 8

66 TABLE 2.5: MODELLED DOMAIN, SELECTED DISCRETE RECEPTORS Receptor Description Grid Reference R1 Residential, southern Buchanan , R2 Residential, southern Buchanan , R3 Residential, southern Buchanan , R4 Residential, southern Buchanan , R5 UNMIL Barracks, NW of proposed power plant , R6 Red Sea accommodation compound , R7 Loop , R8 Loop , R9 Loop , R10 AML accommodation, E of Loop , R11 AML accommodation, E of Loop , R12 Loop , R13 Loop , R14 Loop , R15 Loop , R16 AML accommodation, E of Loop , R17 Residential, southern Buchanan , Note: The x and y coordinates listed are referenced to the WGS co-ordinate system, UTM zone 29N Meteorological Data Project specific meteorological data, collected by AML, has been used in this study. Data collected at the port site itself has been selected for use in the assessment. The measurement site is less than 0.5 km from the proposed power plant. The assessment uses data collected between 1 July 2011 and 30 June 2012, a period of twelve months. The data capture rates have been good over the 12 month period, and has captured the seasonal variation in climatic conditions and prevailing wind direction at the site. Although multiple years of meteorological data would normally be employed in a study such as this one to account for the variation in meteorological conditions from one year to the next, it is considered that the use of local data is of higher importance in this case, rather than a longer period from a location that is less representative of local climatic conditions at the port site. A visual representation of the wind speed and direction data used in the assessment is shown in the wind rose presented in Figure 2.1. The assessment does not use the wind roses to infer the magnitude or frequency of impacts at any receptor. Instead, the hourly sequential observation data are used in the dispersion model to calculate robust estimates of impacts. VOLUME 3, PART 5, APPENDIX 2: BUCHANAN DISPERSION MODEL 9

67 Figure 2.1: Wind Rose for Buchanan Port, 1 July 2011 to 30 June 2012 Building Downwash Effects The buildings that make up the power plant have the potential to affect the dispersion of emissions from the main stack. The ADMS buildings effect module has therefore been used to incorporate building downwash effects as part of the modelling procedure. Buildings greater than approximately one third of the preferred stack height have been included within the modelling assessment. At the current time, the design of the power plant buildings and fuel tanks have not been finalised, so the evaluation of building downwash has been based on preliminary building dimensions. Only the turbine hall is above one third of the proposed stack height of 32 m. The building dimensions, as represented within the model, are presented in Table 2.6. The turbine house has a sloping roof, and has been represented within the model as being the distance from the floor to the apex (including the roof vent). VOLUME 3, PART 5, APPENDIX 2: BUCHANAN DISPERSION MODEL 10

68 TABLE 2.6: BUILDING PARAMETERS Building Grid Reference of Centre Point Length (m) Width (m) Height (m) Angle ( ) Turbine Hall , Note: The x and y coordinates listed are referenced to the WGS co-ordinate system, UTM zone 29N Terrain The land in the area around the power plant site is largely flat, without any sharp changes in gradient and level within the modelled domain. For this reason, the consideration of terrain effects within the model has not been necessary. Surface Roughness A surface roughness of 0.5 m was used within ADMS. This option is considered as representative of the low buildings and sparsely vegetated landscape between the emission point and the receptors. As the meteorological data used was taken at the site itself, the same surface roughness value was used for the meteorological station. NO X to NO 2 Conversion Emissions of NO x from the main stack will consist mainly of nitric oxide (NO) at the point of release, oxidising within the atmosphere to form NO 2 as it moves downwind. At the point of release, approximately 95% of the NO X emission will be in the form of NO, with the other 5% consisting of NO 2. In the assessment of the impact on annual mean concentrations of NO 2, a worst case approach has been taken and a 100% NO x to NO 2 conversion rate at ground level has been assumed in the calculation of long-term annual mean calculations. With regard to short-term impacts, even within such a lightly populated environment, the oxidisation of NO to NO 2 is a relatively slow process, and for this reason it is considered that the assumption of 100% conversion within the modelled domain is excessively conservative in locations close to the power plant stacks. A 50% conversion rate has therefore been assumed in the calculation of short-term hourly concentrations. Specialised Model Treatments Emissions have been modelled such that they are not subject to dry and wet deposition. This assumption of continuity of mass is likely to result in an over-estimation of impacts at distant receptors. VOLUME 3, PART 5, APPENDIX 2: BUCHANAN DISPERSION MODEL 11

69 3 BASELINE CONDITIONS 3.1 Receptor Locations The town of Buchanan lies to the north of the port concession area. There are also a number of residential areas situated to the east and south east of the power plant, including the UNMIL barracks and AML residential properties Measured Dust and Particulate Matter Concentrations A summary of PM 10 and PM 2.5 concentrations measured over the period November 2008 to October 2009 is provided in Table 3.1. Although data capture rates were low due to sitespecific difficulties, the data indicates that period mean PM 10 and PM 2.5 concentrations at Buchanan are within the annual mean assessment criteria. TABLE 3.1: MEASURED PM 10 AND PM 2.5 CONCENTRATIONS Site Data Capture Period Mean PM 10 PM 2.5 Buchanan 54% 66 9 The measured PM 2.5 concentrations are significantly lower than the PM 10 levels. This indicates a significant coarse dust emissions source, consistent with re-suspended dust from unpaved roads Measured Baseline Nitrogen Dioxide and Sulphur Dioxide The baseline diffusion tube survey in the area around Buchanan commenced in March The survey was programmed to run for six months, but site specific difficulties have resulted in some loss of data. At the time of writing, the results of monitoring from March to May 2012 were available for NO 2, and March to July 2012 in respect of SO 2. These results are presented in Table 3.2. Although concentrations measured from month to month will vary due to meteorological conditions and seasonality, the results indicate that, as expected, baseline concentrations of combustion pollutants within the air quality study area are very low in most locations. Higher concentrations of NO 2 were recorded in close proximity to the train unloading and stocking facility at the port, but there is no residential exposure at this location. At all the monitoring sites, the data indicates that there is no existing risk of a breach of the long term and short term air quality standards for NO 2 and SO 2. TABLE 3.2: DIFFUSION TUBE MONITORING RESULTS, MARCH TO JULY 2012 Mean Measured Pollutant Concentration (µg/m 3 ) Site ID Location NO 2 SO 2 B1 AML office B2 Moore Town Gate B3 Port Gate VOLUME 3, PART 5, APPENDIX 2: BUCHANAN DISPERSION MODEL 12

70 4 DISPERSION MODELLING RESULTS This section reports the results of the dispersion modelling of emissions from the port power plant, at the preferred stack height of 32 m, with 2% maximum sulphur fuel. Where it is possible to do so, baseline pollutant concentrations relevant to the receptor in question have been taken from the project specific monitoring data presented in Section 3 of this report. Short term background concentrations (averaging periods within a day) have been assumed to be double the long term concentration. Consideration of Stack Height The modelling study has included an evaluation of the height of the power plant stacks, using the ADMS dispersion model. The selection of an appropriate stack height requires a number of factors to be taken into account, the most important of which is the need to balance a stack height sufficient to achieve adequate dispersion of pollutants against other constraints such as cost and visual impact. Emissions from the main stack have been modelled at stack heights between 20 m and 80 m. A graph, showing the process contribution (PC) to annual mean NO 2 concentrations is presented in Plot 4.1, below. The purpose of the graph is to evaluate the optimum stack height in terms of the dispersion of pollutants which would occur, against the other constraints of further increases in release height. Analysis of the annual mean curve shows that the benefit of incremental increases in stack height up to 30 m is more pronounced. At heights above m, the curve flattens and the air quality benefit of increasing stack height further is reduced. It is therefore considered that 32 m is around the height at which site-specific constraints and the diminishing benefits of further increases in release height begin to outweigh the benefits to air quality. VOLUME 3, PART 5, APPENDIX 2: BUCHANAN DISPERSION MODEL 13

71 Plot 4.1: Predicted Process Contribution to Maximum Annual Mean NO 2 Concentrations at Stack Heights between 20 m and 80 m Annual Mean NO2 PC (µg/m 3 ) Stack (m) Nitrogen Dioxide (NO 2 ) The predicted change in annual mean NO 2 concentrations that would occur during the operation of the power plant, at the selected sensitive receptors, is presented in Table 4.1. The predicted impact on short term NO 2 concentrations is presented in Table 4.2. Pollutant contour plots, showing the spatial distribution of impacts, are presented in Figure 4.1 and Figure 4.2. VOLUME 3, PART 5, APPENDIX 2: BUCHANAN DISPERSION MODEL 14

72 TABLE 4.1: PREDICTED IMPACT ON ANNUAL MEAN NO 2 CONCENTRATIONS Receptor Description Annual Mean NO 2 Concentration (µg/m 3 ) Bkg PC PEC PEC % AQS R1 Residential, southern Buchanan R2 Residential, southern Buchanan R3 Residential, southern Buchanan R4 Residential, southern Buchanan R5 UNMIL Barracks R6 Red Sea accommodation compound R7 Loop R8 Loop R9 Loop R10 AML accommodation, E of Loop R11 AML accommodation, E of Loop R12 Loop R13 Loop R14 Loop R15 Loop R16 AML accommodation, E of Loop R17 Residential, southern Buchanan Air Quality Standard (IFC) 40 The results in Table 4.1 show that the operation of the power plant would not cause a risk of exceeding the annual mean NO 2 air quality standard at the selected sensitive receptors where relevant human exposure would occur. Due to its proximity to the power plant site, the most affected residential location would be the Red Sea accommodation compound and parts of Loop 1, where the predicted environmental concentration (PEC) would remain below 40% of the annual mean standard. As for the hourly mean, the largest impact on short-term NO 2 concentrations would be seen in the vicinity of the Red Sea compound and Loop 1. At this location and the other sensitive receptors considered within the assessment, there would not be a predicted exceedance of the 1-hour standard. VOLUME 3, PART 5, APPENDIX 2: BUCHANAN DISPERSION MODEL 15

73 TABLE 4.2: PREDICTED IMPACT ON MAXIMUM 1 HOUR MEAN NO 2 CONCENTRATIONS Receptor Description 1-hour Mean NO 2 Concentration (µg/m 3 ) Bkg PC PEC PEC % AQS R1 Residential, southern Buchanan R2 Residential, southern Buchanan R3 Residential, southern Buchanan R4 Residential, southern Buchanan R5 UNMIL Barracks R6 Red Sea accommodation compound R7 Loop R8 Loop R9 Loop R10 AML accommodation, E of Loop R11 AML accommodation, E of Loop R12 Loop R13 Loop R14 Loop R15 Loop R16 AML accommodation, E of Loop R17 Residential, southern Buchanan Air Quality Standard (IFC) 200 Sulphur Dioxide (SO 2 ) (2% maximum S content in the fuel) The predicted change in 24-hour mean SO 2 concentrations that would occur during the operation of the power plant, at the selected sensitive receptors with 2% sulphur fuel, is presented in Table 4.3. The predicted impact on 10-minute SO 2 concentrations is presented in Table 4.4. Pollutant contour plots, showing the spatial distribution of impacts, are presented in Figure 4.3 and Figure 4.4. The results show that, if 2% sulphur fuel is used for power generation, there would be a low risk of exceeding the short-term air quality standards for SO 2 at sensitive receptor locations. The predicted maximum 24-hour and 1-hour concentrations are around 40% and 50% of the air quality standards respectively. VOLUME 3, PART 5, APPENDIX 2: BUCHANAN DISPERSION MODEL 16

74 TABLE 4.3: PREDICTED IMPACT ON 24 HOUR MEAN SO 2 CONCENTRATIONS Receptor Description 24-hour Mean SO 2 Concentration (µg/m 3 ) Bkg PC PEC PEC % AQS R1 Residential, southern Buchanan R2 Residential, southern Buchanan R3 Residential, southern Buchanan R4 Residential, southern Buchanan R5 UNMIL Barracks R6 Red Sea accommodation compound R7 Loop R8 Loop R9 Loop R10 AML accommodation, E of Loop R11 AML accommodation, E of Loop R12 Loop R13 Loop R14 Loop R15 Loop R16 AML accommodation, E of Loop R17 Residential, southern Buchanan Air Quality Standard (IFC) 125 VOLUME 3, PART 5, APPENDIX 2: BUCHANAN DISPERSION MODEL 17

75 TABLE 4.4: PREDICTED IMPACT ON MAXIMUM 10 MINUTE MEAN SO 2 CONCENTRATIONS Receptor Description 10-minute Mean SO 2 Concentration (µg/m 3 ) Bkg PC PEC PEC % AQS R1 Residential, southern Buchanan R2 Residential, southern Buchanan R3 Residential, southern Buchanan R4 Residential, southern Buchanan R5 UNMIL Barracks R6 Red Sea accommodation compound R7 Loop R8 Loop R9 Loop R10 AML accommodation, E of Loop R11 AML accommodation, E of Loop R12 Loop R13 Loop R14 Loop R15 Loop R16 AML accommodation, E of Loop R17 Residential, southern Buchanan Air Quality Standard (IFC) 500 Particulate Matter (PM 10 and PM 2.5 ) The predicted change in annual mean PM 10 concentrations that would occur during the operation of the power plant, at the selected sensitive receptors, is presented in Table 4.5. The predicted impact on short term PM 10 concentrations is presented in Table 4.6. The predicted change in annual mean PM 2.5 concentrations that would occur during the operation of the power plant, at the selected sensitive receptors, is presented in Table 4.7. The predicted impact on short term PM 2.5 concentrations is presented in Table 4.8. Pollutant contour plots, showing the spatial distribution of predicted impacts on PM 10 concentrations, are presented in Figure 4.5 and Figure 4.6. As the total particulate matter emission is assumed to be present as both PM 10 and PM 2.5, these contour plots apply to the predicted change in PM 2.5 also. The results tables show that the impact on local concentrations of particulate matter from the operation of the power plant would be very low in comparison to existing baseline concentrations. Although the PM 10 annual mean air quality standards are predicted to be at risk of exceedance, this is overwhelmingly due to existing concentrations and not the predicted impact of the operational power plant. VOLUME 3, PART 5, APPENDIX 2: BUCHANAN DISPERSION MODEL 18

76 TABLE 4.5: PREDICTED IMPACT ON ANNUAL MEAN PM 10 CONCENTRATIONS Receptor Description Annual Mean PM 10 Concentration (µg/m 3 ) Bkg PC PEC PEC % AQS R1 Residential, southern Buchanan R2 Residential, southern Buchanan R3 Residential, southern Buchanan R4 Residential, southern Buchanan R5 UNMIL Barracks R6 Red Sea accommodation compound R7 Loop R8 Loop R9 Loop R10 AML accommodation, E of Loop R11 AML accommodation, E of Loop R12 Loop R13 Loop R14 Loop R15 Loop R16 AML accommodation, E of Loop R17 Residential, southern Buchanan Air Quality Standard (IFC) 70 VOLUME 3, PART 5, APPENDIX 2: BUCHANAN DISPERSION MODEL 19

77 TABLE 4.6: PREDICTED IMPACT ON 99 TH PERCENTILE 24 HOUR MEAN PM 10 CONCENTRATIONS Receptor Description 24-hour Mean PM 10 Concentration (µg/m 3 ) Bkg PC PEC PEC % AQS R1 Residential, southern Buchanan R2 Residential, southern Buchanan R3 Residential, southern Buchanan R4 Residential, southern Buchanan R5 UNMIL Barracks R6 Red Sea accommodation compound R7 Loop R8 Loop R9 Loop R10 AML accommodation, E of Loop R11 AML accommodation, E of Loop R12 Loop R13 Loop R14 Loop R15 Loop R16 AML accommodation, E of Loop R17 Residential, southern Buchanan Air Quality Standard (IFC) 150 VOLUME 3, PART 5, APPENDIX 2: BUCHANAN DISPERSION MODEL 20

78 TABLE 4.7: PREDICTED IMPACT ON ANNUAL MEAN PM 2.5 CONCENTRATIONS Receptor Description Annual Mean PM 2.5 Concentration (µg/m 3 ) Bkg PC PEC PEC % AQS R1 Residential, southern Buchanan R2 Residential, southern Buchanan R3 Residential, southern Buchanan R4 Residential, southern Buchanan R5 UNMIL Barracks R6 Red Sea accommodation compound R7 Loop R8 Loop R9 Loop R10 AML accommodation, E of Loop R11 AML accommodation, E of Loop R12 Loop R13 Loop R14 Loop R15 Loop R16 AML accommodation, E of Loop R17 Residential, southern Buchanan Air Quality Standard (IFC) 35 VOLUME 3, PART 5, APPENDIX 2: BUCHANAN DISPERSION MODEL 21

79 TABLE 4.8: PREDICTED IMPACT ON 99 TH PERCENTILE 24 HOUR MEAN PM 2.5 CONCENTRATIONS Receptor Description 24-hour Mean PM 2.5 Concentration (µg/m 3 ) Bkg PC PEC PEC % AQS R1 Residential, southern Buchanan R2 Residential, southern Buchanan R3 Residential, southern Buchanan R4 Residential, southern Buchanan R5 UNMIL Barracks R6 Red Sea accommodation compound R7 Loop R8 Loop R9 Loop R10 AML accommodation, E of Loop R11 AML accommodation, E of Loop R12 Loop R13 Loop R14 Loop R15 Loop R16 AML accommodation, E of Loop R17 Residential, southern Buchanan Air Quality Standard (IFC) 75 VOLUME 3, PART 5, APPENDIX 2: BUCHANAN DISPERSION MODEL 22

80 Carbon Monoxide (CO) The predicted change in 8-hour mean CO concentrations that would occur during the operation of the power plant, at the selected sensitive receptors, is presented in Table 4.9. Although no baseline data is available, due to the lack of large combustion sources in the area it is very unlikely that the air quality standard would be exceeded. TABLE 4.9: PREDICTED IMPACT ON 8 HOUR MEAN CO CONCENTRATIONS Receptor Description 8-hour Mean CO Concentration (µg/m 3 ) Bkg PC PEC PC % AQS R1 Residential, southern Buchanan <1 R2 Residential, southern Buchanan <1 R3 Residential, southern Buchanan <1 R4 Residential, southern Buchanan <1 R5 UNMIL Barracks <1 R6 Red Sea accommodation compound <1 R7 Loop <1 R8 Loop <1 R9 Loop <1 R10 AML accommodation, E of Loop <1 R11 AML accommodation, E of Loop <1 R12 Loop <1 R13 Loop <1 R14 Loop <1 R15 Loop <1 R16 AML accommodation, E of Loop <1 R17 Residential, southern Buchanan <1 Air Quality Standard (IFC) 10,000 VOLUME 3, PART 5, APPENDIX 2: BUCHANAN DISPERSION MODEL 23

81 Maximum Impacts Within the Modelled Domain The maximum Process Contribution (PC) and Predicted Environmental Concentration (PEC) within the modelled domain, including locations where there is no relevant human exposure, are summarised in Table 4.10 for each pollutant and averaging period. The distribution of impacts is also illustrated in Figure 4.1 to Figure 4.6 inclusive. TABLE 4.10: 32M STACK HEIGHT WITH 2% SULPHUR FUEL - MAXIMUM PROCESS CONTRIBUTION AND PREDICTED ENVIRONMENTAL CONCENTRATION, ALL MODELLED POLLUTANTS Pollutant Averaging Period AQS PC PC PEC PEC (µg/m 3 ) (µg/m 3 ) (% EAL) (µg/m 3 ) (% EAL) Annual Mean NO 2 Maximum 1-hour mean Annual Mean PM th %ile of 24-hour means Annual Mean PM 2.5 SO 2 CO 99 th %ile of 24-hour means Maximum 24-hour mean Maximum 10-min mean Max daily 8-hour running mean , <0.1 n/a n/a It can be seen that the operation of the power plant would not cause an exceedance of the air quality standards with 2% sulphur fuel and a 32 m stack, at any location within the modelled domain. With a 32m stack, the short term NO 2 standard would be at risk of being exceeded in a small area to the west of the power plant, within the port development boundary. As members of the public would not be present at these locations, however, it is more relevant to consider impacts at sensitive receptor locations where human exposure may occur. The PC to PM 10 and PM 2.5 concentrations would not directly cause an exceedance of the air quality standards at any location within the modelled domain, although high baseline concentrations at some locations would mean that the annual mean air quality standard for this pollutant would be at risk of exceedance both with and without the power plant in operation. IFC General EHS guidelines state that to allow for future sustainable development within the same airshed, emissions from a single project should not normally contribute more than 25% of a relevant air quality standard. In this case, however, it is considered that there would not be a significant chance of extensive further industrial development in the immediate vicinity of the area immediately around the power plant over the life of the project. Therefore, the proposed development does not have the potential to restrict further sustainable development in the area. VOLUME 3, PART 5, APPENDIX 2: BUCHANAN DISPERSION MODEL 24

82 Consideration of Power Plant Operation with 4% Sulphur Fuel In addition to the consideration of SO 2 impacts from the power plant operating with fuel containing the IFC guideline limit of 2% sulphur, the dispersion modelling assessment has also evaluated the impact of the use of fuel containing 4% sulphur on stack height. Table 4.11 presents a comparison of maximum predicted 24-hour PC values for a range of stack heights at the most affected sensitive receptor (an area including Loop 1 and the Red Sea accommodation compound). The results are reported both as the PC in µg/m 3 and as a percentage of the IFC guideline air quality standard. A comparison of predicted impacts on maximum 10-minute concentrations is presented in Table TABLE 4.11: COMPARISON OF PREDICTED IMPACTS ON MAXIMUM 24 HOUR MEAN SO 2 CONCENTRATIONS AT THE MOST AFFECTED SENSITIVE RECEPTOR Stack Height PC with 2% S 24-hour Mean SO 2 Concentration (µg/m 3 ) % AQS with 2% S PC with 4% S % AQS with 4% S 20 m m m m m m m m m m m m m m m m VOLUME 3, PART 5, APPENDIX 2: BUCHANAN DISPERSION MODEL 25

83 TABLE 4.12: COMPARISON OF PREDICTED IMPACTS ON MAXIMUM 10 MINUTE MEAN SO 2 CONCENTRATIONS IN THE NORTHERN SECTION OF LOOP 1 Stack Height PC with 2% S 10 Minute Mean SO 2 Concentration (µg/m 3 ) % AQS with 2% S PC with 4% S % AQS with 4% S 20 m m m m m m m m m m m m m m m m Table 4.11 indicates that the 24-hour standard is unlikely to be exceeded with either fuel at Loop 1, with a 32 m stack. The PC to 10-minute mean concentrations, however, show that the predicted PC would be more than 90% of the standard with 4% sulphur fuel at stack heights up to approximately 33 m. With a 35 m stack, the PC decreases to around 75% of the 10- minute air quality standard for 4% sulphur fuel. If supplies of HFO with sulphur contents of less than 2% cannot be guaranteed by the supplier, then it is recommended that the stack height is set within the range of 33 m to 35 m above ground level. VOLUME 3, PART 5, APPENDIX 2: BUCHANAN DISPERSION MODEL 26

84 5 CONCLUSIONS Dispersion modelling of emissions from the Buchanan power plant has demonstrated that a stack height of 32 m would be sufficient to ensure there would not be a breach of ambient air quality standards at sensitive receptor locations, if the plant is operated within emissions limits specified by IFC guidance. If the sulphur content of the HFO fuel cannot be guaranteed by the supplier to be below the IFC standard of 2%, then it is recommended that the stack height is set within the range of 33 m to 35 m above ground level. VOLUME 3, PART 5, APPENDIX 2: BUCHANAN DISPERSION MODEL 27

85 6 REFERENCES CERC, 2007, ADMS 4 Validation Papers IFC (2007) Environmental, Health, and Safety General Guidelines, Washington DC: IFC/World Bank Group IFC (2007) Environmental, Health, and Safety Guidelines for Thermal Power Plants, Washington DC: IFC/World Bank Group WHO (2000). Air Quality Guidelines for Europe. 2 nd ed. (European Series: No. 91). Copenhagen: World Health Organisation Regional Office for Europe. WHO (2005). Air Quality Guidelines Global Update Report on a Working Group Meeting, Bonn, Germany, October 2005 VOLUME 3, PART 5, APPENDIX 2: BUCHANAN DISPERSION MODEL 28

86 7 APPENDIX: FIGURES VOLUME 3, PART 5, APPENDIX 2: BUCHANAN DISPERSION MODEL 29

87 VOLUME 3, PART 5.2: BUCHANAN POWER PLANT DISPERSION MODELLING ASSESSMENT 1

88 VOLUME 3, PART 5.2: BUCHANAN POWER PLANT DISPERSION MODELLING ASSESSMENT 2

89 VOLUME 3, PART 5.2: BUCHANAN POWER PLANT DISPERSION MODELLING ASSESSMENT 3

90 VOLUME 3, PART 5.2: BUCHANAN POWER PLANT DISPERSION MODELLING ASSESSMENT 4