REPORT. Soil, Land Capability and Land Use Assessment Report for the Proposed New Vaal Colliery Life Extension Project.

Rehab Green cc Registration No: 2002/094339/23 PI Steenekamp PO Box 12636 Queenswood 0121 Pretoria Cell: 082 560 0592 Fax: 086 678 1690 E-mail: rehabgreen@ee-sa.com Report No: RG/2012/03/09/1 Date: 30 May 2012 REPORT Soil, Land Capability and Land Use Assessment Report for the Proposed New Vaal Colliery Life Extension Project Requested By Golder Associates Africa (Pty) Ltd Compiled By Rehab Green Monitoring Consultants CC Environmental and Rehabilitation Monitoring Consultant cc P.I. Steenekamp (Cert.Sci.Nat.) Declaration of Independence In terms of Section 33 of the EIA Regulations 2006 published in terms of Chapter 5 of the National Environmental Management Act (Act 107 of 1998) specialists involved in Impact Assessment processes must declare their independence and furnish details of experience. I, Piet Steenekamp, hereby declare that I have no conflict of interest related to the work of this report. Specially, I declare that I have no personal financial interests in the property and/or development being assessed in this report, and that I have no personal or financial connections to the relevant property owners, developers, planners, financiers or consultants of the development. I declare that the opinions expressed in this report are my own and a true reflection of my professional expertise. P.I. Steenekamp Date: 17 May 2012

2 CONTENTS Page 1. INTRODUCTION 5 1.1 Project background 5 1.2 Scope of work 6 2. STUDY OBJECTIVES 6 3. STUDY AREA 7 3.1 Proposed opencast, underground and infrastructure areas 7 4. METHODOLOGY 9 4.1 Preparation of field maps 9 4.2 Soil classification 9 4.3 Soil sampling and analyses 9 4.4 Land capability assessment 10 4.5 Agricultural potential classification 10 4.6 Wetland and riparian delineation 10 4.7 Land use mapping 11 4.8 Erodibility evaluation 11 4.9 Map compilations 11 4.10 Impact Assessment 11 5. SURVEY RESULTS 13 5.1 Dominant soil types 13 5.1.1 Soil types within New Cornelia Block 1 opencast and infrastructure area 14 5.1.2 Soil types within the New Cornelia Vaalbank underground mining area and conveyer route 18 5.2 Other soil qualities derived from soil properties 24 5.3 Soil chemistry 24 5.4 Land capability and agricultural potential 27 5.4.1 Land capability of New Cornelia Block 1 opencast and infrastructure area 28 5.4.2 Land capability of New Cornelia Vaalbank underground mining area and conveyer route 30 5.5 Wetland and riparian delineation 32 5.6 Pre-mining land use 33 5.6.1 Pre-mining land use within the New Cornelia Block 1 opencast and infrastructure areas 33 5.6.2 Pre-mining land use within the New Cornelia Vaalbank underground mining area and conveyer route 34

3 5.7 Historical agricultural production 36 5.8 Evidence of misuse 36 6. PROCEDURES FOR STRIPPING AND HANDLING OF SOIL 37 6.1 Principles in stripping and handling of topsoil 37 6.2 Construction phase 38 6.3 Operational phase (Open Pit) 38 6.3.1 Post-mining land capability commitments 41 6.4 Decommissioning phase 42 7. ENVIRONMENTAL IMPACT ASSESSMENT 42 7.1 Construction phase 42 7.2 Operational phase 42 7.3 Decommissioning phase 42 8. CONCLUSIONS 50 8.1 Soils and land capability 50 8.1.1 Proposed New Cornelia Block 1 opencast area 50 8.1.2 Proposed New Cornelia Vaalbank underground mining area and Conveyer route 50 8.2 Pre-mining land use 51 8.2.1 Proposed New Cornelia Block 1 opencast area 51 8.2.2 Proposed New Cornelia Vaalbank underground mining area and Conveyer route 51 8.3 Impact assessment 51 9. RECOMMENDATIONS 52 9.1 Requirements for successful rehabilitation 52 REFERENCES TABLES: Table 1: Impact assessment methodology 12 Table 2a Soil legend New Cornelia Block 1 opencast and infrastructure area 17 Table 2b Soil legend Cornelia Vaalbank 20 Table 3 Other derived soil properties 24 Table 4 Soil chemical analyses 25 Table 5 Soil fertility compared to broad fertility guidelines 25 Table 6a Land capability classes, soils and agricultural potential New Cornelia Block 1 opencast and infrastructure areas 29 Table 6b Land capability classes, soils and agricultural potential New Cornelia Vaalbank underground mining area and conveyer route 31 Table 7a Pre-mining land uses within the New Cornelia Block 1 opencast and

4 Infrastructure areas 34 Table 7b Pre-mining land uses within the New Cornelia Vaalbank underground Mining area and conveyer route 34 Table 8a Historical agricultural production New Cornelia Block 1 36 Table 8b Historical agricultural production New Cornelia Vaalbank 36 Table 9 Criteria for soil stripping and replacing at New Cornelia Block 1 opencast 40 Table 10 Impact assessment Construction phase 43 Table 11 Impact assessment Operational phase 45 FIGURES: Figure 1 Regional setting of Newvaal Lifex proposed mining areas 5 Figure 2 Study area and key components of the proposed operation 8 Figure 3 a Soil map of the proposed New Cornelia Block 1 open pit and Infrastructure areas 14 Figure 3b Soil map of New Cornelia Vaalbank underground mining area 18 Figure 4 a Land capability map of New Cornelia Block 1 28 Figure 4b Land capability map of the New Cornelia Vaalbank 30 Figure 5a Land use map of New Cornelia Block 1 opencast and infrastructure Areas 33 Figure 5b Land use map of New Cornelia Vaalbank underground mining area And conveyer route 35 Figure 6 Soil stripping and stockpile guide for New Cornelia Block 1 open pit 41 Appendices Appendix A Soil classification system 54 Appendix B Soil properties and characteristics 55 Appendix C Wetland delineation 57 Appendix D Coordinates of soil sampling and stake points 60 Appendix E Soil horizons properties influencing stripping and stockpiling procedures 62

5 1. INTRODUCTION 1.1 Project background The New Vaal Colliery (NVC) Life Extension (Lifex) Project is generally referred to as the NVC Lifex Project. Anglo American Thermal Coal s (AATC) New Vaal Colliery in the Free State Province is situated approximately 7 km east of the town of Vanderbijlpark and immediately south of Vereeniging (Figure 1). Eskom needs a secure coal supply to keep the Lethabo Power Station running until 2043. Anglo American Thermal Coal (AATC) can fulfill this requirement if the life of New Vaal Colliery (NVC) is extended by 13 years, supplemented with additional coal sourced from resources further south of NVC. AATC proposes therefore to blend coal from NVC with coal from the proposed new operations to the south of the existing mine. These operations will include both opencast and underground mining within the following mining blocks (Figure 1): New Cornelia Block 1 (opencast); and New Cornelia Vaalbank (underground). Figure 1: Regional setting of NVC Lifex proposed mining areas

6 1.2 Scope of work Rehab Green Monitoring Consultants cc was requested by Golder Associates Africa (Pty) Ltd (Golder) to conduct a detailed soil, land capability and land use assessment of the proposed New Cornelia Block 1 opencast area as well as higher level field assessment of the New Cornelia Vaalbank proposed underground mining areas. The New Cornelia Block 1 proposed opencast area was surveyed during October 2010 and January 2011 and the survey results were compiled in Rehab Green cc Report No. RG/2010/07/10/1, June 2011. The proposed underground mining area of New Cornelia, as well as a 500 m wide proposed conveyer corridor was surveyed during January and February 2011 and the survey results were compiled in Rehab Green cc Report No. RG/2010/07/10/1/2, April 2011. Rehab Green cc was requested to combine the soil information of the New Cornelia Block 1 opencast and Vaalbank underground mining areas into one report and include new sections for a repositioned conveyer, a new power line, new haul road section, a brine pond, water balancing dam and a water treatment plant. The original New Cornelia Block 1 opencast area was refined and a strip along the western edge was excluded. The soil information was amended based on the refined open pit area. The study provides input to the Environmental Impact Assessment (EIA) as required in terms of the Mineral and Petroleum Resources Development Act (MPRDA), Act 28 of 2002 and the National Environmental Management Act (NEMA), Act 107 of 1998. The Acts require that pollution and/or degradation of the environment is to be avoided, or where either aspect cannot be avoided, to be minimized and remedied. 2. STUDY OBJECTIVES The study objectives were to: Conduct as detailed soil assessment of the proposed opencast mining areas; Conduct a higher level soil assessment of the proposed underground mining and infrastructure areas; Classify and map soil forms according to the South African Taxonomic Soil Classification System, 1991; Derive and map land capability based on soil properties; Map all pre-mining and current land uses; Identify soil properties related to wetness to enable the delineation of zones with similar degrees of wetness categorized in permanent, seasonal or temporary wetland zones or riparian zones based on guidelines of the Department of Water Affairs (DWA); Locate a 100 m buffer line along the outer edge of wetland and riparian zones; Determine all possible impacts by the proposed operations and provide associated mitigation measures; and Compile a soil stripping and stockpiling plan with rehabilitation guidelines.

7 3. STUDY AREA 3.1 Proposed opencast, underground and infrastructure areas The originally surveyed New Cornelia Block 1 opencast area, demarcated in pink, comprises 874.1 ha (Figure 2). An amended open pit area, demarcated with a yellow dotted line, was received from Golder, comprising 648 ha. The amended open pit area excludes a large portion of the wetland on the western edge of the original surveyed area. This section was excluded on all the amended soil, land capability and land use maps in this report. The New Cornelia Block 1 study area was expanded to the north to accommodate a proposed brine pond (5.5 ha), balancing dam (11.7 ha) and water treatment facility (0.8 ha), demarcated in turquoise. The study area also includes the proposed new haul road and ramps, shown in purple in Figure 2. The New Cornelia Vaalbank underground mining area, comprising 4820 ha, demarcated in red, was expanded to include the proposed new power line (green) stretching from the existing Makalu substation to the proposed new shaft area (hatched in black). A new conveyer and power line (pink and black respectively) runs parallel from the new shaft area to the existing Lethabo power station. The New Cornelia Block 1 study area is situated on portions of the farms Rietfontein 152, Zandfontein 259, Driefontein 2 and Waagstuk 78. The study area is bordered by the R59 highway and Vaal Race Course to the north, a railway line to the east and the Taaibosspruit to the west. The study area consist of flat to gentle footslopes to the east and valley bottoms to the west and the area is utilised for cattle grazing and sand mining operations. The New Cornelia Vaalbank study area consists of portions of the farms Modderfontein 1311, Zuurbekom 1615, Modderfontein 1279, Smalstreep 1310, Kruissementfontein 84, Die Hoop 1636, Khadie 1309, Emden 1312, Vaalbank 238, Vaalbank 219, Verdun 1138 and Amelia 518. The study area is roughly bordered by the R716 to the north east, the rail road and R57 to the west and southwest and the R549 to the south. The Taaibosstpruit and several of its tributaries cuts through the study area. The major land uses are the cultivation of maize and soybeans (40%, including fallow lands), pastures (10%) and grazing (47%). The majority of the study area consists of flat to gently sloped crests, midslopes and footslopes with average slopes of 0-4%.

8

9 4. METHODOLOGY 4.1 Preparation of field maps ArcGIS Geographic Information System (GIS) software was used to process all available data for accurate surveying. The location and extents of the mining application areas and infrastructure were obtained from Golder in electronic shapefile format. The shapefiles were superimposed on a Google Earth image as well as 1:50 000 scale topographic sheets. A grid of field observation points were generated at a density of 150 x 150 m. The coordinates of the observation points were calculated and loaded on a Geographic Positioning System (GPS) to accurately locate the position of the observation points in the field. Large scale field maps (1:5000 scale) showing the proposed mining area and observation points on both aerial and topographic background data were printed to use during the field assessment. 4.2 Soil classification The soils at New Cornelia Block 1 were investigated by means of auger holes at a density of 150 x 150 m and 300 x 300 m at New Cornelia Vaalbank. Auger holes were made to a depth of 1500 mm or to the depth of refusal. The soils were described and classified according to the South African Taxonomic Soil Classification System (Soil Classification Working Group, 2nd edition 1991). At New Cornelia Block 1 a total of 331 auger observations were made at grid points and 19 randomly where deemed necessary to locate and accurately map soil boundaries. At New Cornelia Vaalbank a total of 350 auger observations were made at a density of approximately 300 x 300 m except at the farms Khadi 1309 and Emden 1312 where access to the property was denied. The system of soil classification is explained in Appendix A. The following procedure was followed to note soil properties and classify soils accordingly: i) Identify applicable diagnostic horizons by noting the physical properties such as: Effective depth (depth of soil suitable for root development); Colour (in accordance with Munsell colour chart); Texture (refers to the particle size distribution); Structure (aggregation of soil particles into structural units); Mottling (alterations due to continued exposure to wetness); Concretions (cohesion of minerals into hard fragments); Leaching (removal of soluble constituents by percolating water); Gleying ( reduction of ferric oxides under anaerobic conditions resulting in grey, low chroma soil colours); and Illuviation of colloidal mater from one horizon to another resulting in the development of grey sandy E-horizons and grey clay G-horizons. ii) Determine according to above properties the appropriate soil Form and soil Family The soil properties that were used in the soil classification are discussed in Appendix B. 4.3 Soil sampling and analyses The A and B or E or G-horizons (0-250 and 300-700mm) of the dominant soil types were sampled and analysed at the Institute for Soil, Climate and Water. The analyses

10 were conducted according to methods set out in the Handbook of Standard Testing for Advisory Purposes (Soil Science Society of South Africa, 1990). The following analyses were conducted: Soil acidity (ph) in a 1:2.5 water solution; Extractable cations (Na, K, Ca and Mg) according to the ammonium acetate method; and Phosphorus status according to the Bray 1 method. 4.4 Land capability assessment Land capability was assessed according to the definitions outlined in the guidelines for the rehabilitation of mined land by the Chamber of Mines of South Africa and Coaltech Research Association (2007). Soils types were classified into the following categories for areas that exclude wetlands: Arable land; Grazing land; and Wilderness. 4.5 Agricultural potential classification The classification of agricultural potential of soils was based on soil properties noted during auger observations such as effective soil depth, texture, terrain unit, slope, soil wetness and disturbances. The effective soil depth and soil texture is the main soil characteristic that determined the agricultural potential. The criteria applied for the classification of the agricultural potential of soils are as follows: High well-drained and moderately well-drained soils with an effective depth deeper than 900 mm. Moderate - well-drained and moderately well-drained soils with an effective depth of 600-900 mm. Low - well-drained and moderately well-drained soils with an effective depth less than 600 mm or leached, grey, sandy soils showing evidence of periodic percolating water tables, or black and grey clay soils showing evidence of poor internal drainage or very sandy soils in low rainfall areas. 4.6 Wetland and riparian delineation Wetland and riparian zones are delineated according to the practical field procedure for the identification and delineation of wetlands and riparian areas (Department of Water Affair and Forestry, 2005). Four indicators were used in the study to delineate wetland and riparian zones, namely: Terrain unit; Soil form; Soil wetness; and Wetland and riparian vegetation. Auger observations are made systematically in transects towards the wetland at 50 m intervals in order to locate the point from where clear evidence of wetness occurs within 500 mm from surface. Further details on the delineation of wetland areas are included in Appendix C.

11 4.7 Land use mapping The localities and extents of land use practices were surveyed during the time of the soil assessment. 4.8 Erodibility evaluation Erodiblity was broadly assessed based on soil texture, slope and the inherent stability of the parent rock (geology) from which the soil originated. Low: Soils with stable physical and chemical properties which occur on flat to gentle slopes to ensure low erosion susceptibility in the natural state. Few erosion protection measures are necessary. Moderate: Soils with low to moderately unstable physical or chemical properties or soils occurring on moderate to steep slopes. Sheet and rill erosion often occur in the natural state but may become severe when these soils are disturbed or due to any misuse such as overgrazing. Erosion protection measures are necessary. High: Soils with unstable physical and chemical properties or soils occurring on very steep slopes. Rill and donga erosion often occur in the natural state and will become severe during any disturbance or misuse. Specialised erosion protection measures are necessary. 4.9 Map compilations The field data was captured in shapefile format (shp) and processed and stored in a Geographic Information System called ArcGIS, version 9.3.1. The maps are compiled on aerial photo back ground in a map extendable document format (mxd) and exported to Jpeg format. The shapefiles can be exported to a dxf or dwg format for CAD users. The shapefiles, dxf and dwg formats are available on request. The maps were generated in a projected coordinate system using the longitude of origin (LO) coordinate system based on the 27 East meridian, WG1984 Elipsoid and Hartebeesthoek 1994 Datum. 4.10 Impact assessment This assessment evaluates the effects of the proposed project on the soil environment. Each potential impact was assessed according to the following criteria: Magnitude is a measure of the degree of change in a measurement or analysis which is classified as minor/negligible, low, moderate, high or very high. Scale/Geographic extent refers to the area that could be affected by the impact and is classified as none, site only, local, regional, national, or international. Duration refers to the length of time over which an environmental impact may occur: i.e. Immediate (less than 1 year), short-term (0 to 7 years), medium term (8 to 15 years), long-term (greater than 15 years with impact ceasing after closure of the project) or permanent. Probability of occurrence is a description of the probability of the impact actually occurring as improbable (less than 5 % chance), low probability (5 %

12 to 40 % chance), medium probability (40 % to 60 % chance), highly probable (most likely, 60 % to 90 % chance) or definite (impact will definitely occur). Direction of an impact may be positive, neutral or negative with respect to the particular impact. Reversibility is an indicator of the potential for recovery of the endpoint from the impact. Frequency describes how often the impact may occur within a given time period and is classified as low, medium or high frequency. Seasonal considerations should be discussed where these are important in the evaluation of the impact. The significance of the identified impacts was determined using the approach summarized in Table 1. This incorporates two aspects for assessing the potential significance of impacts (terminology from the Department of Environmental Affairs and Tourism Guideline document on EIA Regulations, April 1998), namely occurrence and severity, which are further sub-divided as shown in Table 1. Table 1: Impact assessment methodology OCCURRENCE Probability of occurrence Duration of occurrence To assess each impact, the following four ranking scales are used: Severity Magnitude (severity) of Scale / extent of impact impact PROBABILITY DURATION 5 - Definite/don t know 5 - Permanent 4 - Highly probable 4 - Long-term 3 - Medium probability 3 - Medium-term (8-15 years) 2 - Low probability 1 - Improbable 1 Immediate 0 - None SCALE 2 - Short-term (0-7 years) (impact ceases after the operational life of the activity) 5 - International 10 - Very high/don t know 4 - National 8 - High 3 - Regional 6 - Moderate 2 - Local 4 - Low 1 - Site only 2 - Minor 0 - None MAGNITUDE The significance of the two aspects, occurrence and severity, is assessed using the following formula: SP (significance points) = (magnitude + duration + scale) x probability The maximum value is 150 significance points (SP). The impact significance points are assigned a rating of high, medium or low with respect to their environmental impact as follows: SP >75 SP 30 75 SP <30 Indicates high environmental significance Indicates moderate environmental significance Indicates low environmental significance + Positive impact An impact which could influence the decision about whether or not to proceed with the project regardless of any possible mitigation. An impact or benefit which is sufficiently important to require management and which could have an influence on the decision unless it is mitigated. Impacts with little real effect and which should not have an influence on or require modification of the project design. An impact that is likely to result in positive consequences/effects. Potential impacts were assessed using the above calculation and rating system, and mitigation measures were proposed for all relevant project phases (construction to decommissioning).

13 5. SURVEY RESULTS 5.1 Dominant soil types Soil types were mapped based on soil information gathered by means of auger observations. The soil information of New Cornelia Block 1 and Vaalbank mining areas were compiled on two separate maps, Figures 3a and 3b respectively. The information of soils along the proposed conveyer route is included in the New Cornelia Vaalbank maps and Tables. Each figure contains an abbreviated soil legend. A detailed soil legend for each of the two areas is provided in Tables 2a and 2b which described the soils in terms of the following aspects: Dominant and subdominant soil forms and families; Average effective soil depth; The estimated clay content of the A and B or E or G-horizons; The derived texture class; A description of the terrain unit and slope, a broad description of the dominant soil form; A description of the soil horizon sequences; The soil s derived land capability and wetland zone classification; and The area and percentage comprised by each soil type. 5.1.1 Soil types within the New Cornelia Block 1 opencast and infrastructure areas Soil types within the proposed New Cornelia Block 1 opencast and infrastructure areas are shown on Figure 3a. A total of 5 soil types, based on dominant soil form, effective soil depth, internal drainage, terrain unit and slope percentage were identified during field observations and were symbolised as: Cv, Av, Lo2, Lo1 and Kd/Ka. The soil types are summarised in the soil legend Table 2a.

14

15 Soil type Cv, dominated by the Clovelly soil Form, covering 155.24 ha (18.39% of the study area), occurs on a gentle footslope (1-2% slopes) and consists of very deep, brownish yellow, well-drained, sandy soils. The soil profile sequence consists of yellow brown, sandy, Orthic A-horizons underlain by brownish yellow, sandy B-horizons often underlain by a greyish yellow E-horizon. This non-diagnostic E-horizon indicates some higher leaching in the lower soil profile but mostly lacks any mottling. However, where mottling does occur, the soils were classified as the Pinedene soil form. The higher leaching in the lower soil profile negatively affects the soil potential and land capability due to the inability to effectively store water in the lower soil profile. The general highly leached status of soil type Cv is confirmed by the fairly rapid fading of the prominent yellow colour of the B-Horizon to a more paler yellow when exposed to sunlight. The land capability was classified as grazing potential with low agricultural potential due to the very sandy texture (<5% clay) which manifest in very low water holding capacity and highly leached status. The erodibility was rated low due to stable physical and chemical properties and gentle slopes. Soil type Av, dominated by the Avalon soil Form, covering 424.72 ha (50.32% of the study area), occurs on gentle midslopes (1-2% slopes) and consists of deep, brownish yellow, moderately well-drained, sandy soils. The soil profile sequence consists of yellow brown, sandy, Orthic A-horizons underlain by brownish yellow, sandy B1- horizons underlain by slightly mottled soft plinthic B2-horizons. The mottled soft plinthic B2-horizon often contains a small percentage of soft sesquioxide concretions which is the result of accumulation of iron oxides. A slight increase in the clay content of the B2- horizon will have a positive effect on the water holding capacity in the lower soil profile. A non-diagnostic, greyish yellow E-horizon often occurs between the B1 and B2- horizons. The general highly leached status of soil type Av is confirmed by the fairly rapid fading of the prominent yellow colour of the B1-horizon to a paler yellow when exposed to sunlight. The land capability was classified as grazing potential with low agricultural potential due to the very sandy texture (<5% clay) which manifest in very low water holding capacity and highly leached status. The erodibility was rated low due to stable physical and chemical properties and gentle slopes. Soil type Lo2, dominated by the Longlands 2000 soil Form, covering 133.22 ha (15.78% of the study area), occurs on gentle midslopes (1-2% slopes) and consists of deep, greyish yellow, moderately well-drained, sandy soils. The soil profile sequence consists of greyish brown, sandy, Orthic A-horizons underlain by greyish yellow, sandy E-horizons underlain by slightly mottled soft plinthic B1-horizons. Soil type Lo2 are somewhat similar to soil type Av but are more subjected to leaching. The often fairly strong yellow soil colour below the A-horizon fades to a greyish yellow colour when exposed to sunlight which qualifies as the yellow E-horizon. Although the greyish yellow colour is an indication of wetness it was observed that wetland vegetation constantly lacks in these soils. It is assumed that the degree of wetness in these soils is insufficient to support wetland vegetation. The land capability was classified as grazing potential with low agricultural potential due to the very sandy texture (<5% clay) which manifest in very low water holding capacity and highly leached status. The erodibility was rated low due to stable physical and chemical properties and gentle slopes. Soil type Lo1, dominated by the Longlands soil Form, covering 70.90 ha (8.4% of the study area), occurs on gentle footslopes (1-2% slopes) and consists of deep, greyish yellow to grey, imperfectly drained, sandy soils. The soil profile sequence consists of greyish brown, sandy, Orthic A-horizons underlain by greyish yellow to grey, sandy E- horizons underlain by mottled soft plinthic B1-horizons. Soil type Lo1 is somewhat similar to soil type Lo2 but clearly differ in the degree of wetness in the soil profile. The grey coloured A-horizons indicate wetness in the upper part of the soil profile and the greyish yellow E-horizons and soft plinthic B-horizons indicate periodic wetness in the

16 remainder of the soil profile. Occasional saturated spots with shallow surface water occurred. Water tables were occasionally found in the lower soil profile and wetland vegetation occurred continuously. The land capability was classified as a temporary wetland zone with low agricultural potential due to the degree of wetness in the soil profile. The erodibility was rated low due to stable physical and chemical properties and gentle slopes. Soil type Kd/Ka, dominated by an association of the Kroonstad and Katspruit soil Forms, covering 59.95 ha (7.1% of the study area), occurs on gentle footslopes (1-2% slopes) and consists of shallow, grey, imperfectly to somewhat poorly drained, sandy soils underlain by gleyed clay. The soil profile sequence consists of grey, sandy, Orthic A-horizons underlain by grey to yellowish grey, sandy E-horizons underlain by gleyed clay (G-horizon). A major difference between soil types Kd/Ka and Lo1 is the degree in wetness indicated by the underlying material (G-horizon versus soft plinthic B-horizon). The soft plinthic B-horizon is an indication of periodic wetness while the G-horizon indicates prolonged wetness. The Katspruit soil form, other than the Kroonstad form lacks an E-horizon and consists of a grey, sandy A-horizon directly underlain by a gleyed G-horizon. However, the A-horizons have similar sandy textures and the Katspruit can in this occasion be seen as a shallow version of the kroonstad form. Frequent saturated spots with shallow surface water occurred. Frequent water tables were found and prominent wetland vegetation occurred. The land capability was classified as a seasonal wetland zone with low agricultural potential due to the degree of wetness in the soil profile. The erodibility was rated low due to stable physical and chemical properties and gentle slopes.

17 Table 2a: Soil legend of New Cornelia Block 1 opencast and infrastructure area SOIL LEGEND Soil Type Code Dominant Soil Form and Family Subdominant Soil Form and Family Effective Soil Depth (mm) Cv Clovelly 1100 Pinedene 1100 1200-1600 Av Avalon 1100 Lo2 Longlands 2000 Lo1 Longlands 2000 Kd/Ka Kroonstad 1000 Katspruit 1000 Pinedene 1100 Longlands 2000 Avalon 1100, Pinedene Fernwood 1210, Constantia 1100 Longlands 1000 800-1200 800-1200 500-800 Longlands 1000 200-600 Clay content per horizon A, E, G, B A: 3-5 B: 3-7 E: 1-3 A: 3-5 B1: 3-7 B2: 7-10 A: 3-5 E1: 1-3 B2: 8-15 A: 3-5 E: 1-3 B: 8-20 A: 3-5 E: 1-3 G: 40-55 Sand Sand Sand Sand Texture Class Sandy - clay Summarized Description of Dominant Soil Form Terrain: Gentle footslope (1-2% slopes); Soil: Very deep, brownish yellow, well-drained, sandy soils; Profile: Yellow brown, sandy, Orthic A- horizons underlain by brownish yellow, sandy B- horizons underlain by greyish yellow E-horizons. Terrain: Gentle midslope (1-2% slopes); Soil: Deep, brownish yellow, moderately well-drained, sandy soils; Profile: Yellow brown, sandy, Orthic A-horizons underlain by brownish yellow, sandy B-horizons underlain by slightly mottled soft plinthite. Terrain: Gentle midslope (1-2% slopes); Soil: Deep, greyish yellow, moderately well-drained, sandy soils; Profile: Yellowish brown, sandy, Orthic A-horizons underlain by greyish yellow, sandy E-horizons underlain by slightly mottled soft plinthite. Terrain: Gentle footslope (1-2% slopes); Soil: Deep, greyish yellow to grey, imperfectly drained, sandy soils; Profile: Greyish brown, sandy, Orthic A-horizons underlain by greyish yellow to grey, sandy E-horizons underlain by mottled soft plinthite. Terrain: Gentle footslope (1-2% slopes); Soil: Shallow, grey, imperfectly to somewhat poorly drained, sandy soils; Profile: Grey, sandy, Orthic A-horizons underlain by grey to yellowish grey, sandy E-horizons underlain by gleyed clay (Ghorizon). Agricultu ral Potential Land Capability Erodibility Number of Units Area (ha) Area (%) Low Grazing Low 2 155.24 18.39 Low Grazing Low 1 424.72 50.32 Low Grazing Low 4 133.22 15.78 Low Wetland Low 3 70.90 8.40 Low Wetland Low 3 59.95 7.10 TOTAL 13 844.03 100.0

18 5.1.2 Soil types within the New Cornelia Vaalbank underground mining area and conveyer route Soil types within the proposed New Cornelia underground mining area and conveyer route are shown on Figure 3b. A total of 23 soil types, based on dominant soil form, effective soil depth, internal drainage, terrain unit and slope percentage were identified during field observations and were symbolised as: Hu, Bv, Av1, Av2, We1, We3, Pn- We, Ar, ArR, My-Bo, Ms-Sw, Lo-Kd, Lo1, Lo2, Kd, Se-Ka, Ka, Wo, Rg2, Rg3, Un, Wb and Q. The soil types are summarised in the soil legend Table 2b.

19 Soil types Hu and Bv are deep, red, well- and moderately well-drained soils classified as arable land with high agricultural potential. Soil types Av1 and Av2 are yellow brown, moderately well-drained soils classified as arable land with moderate to high and moderate to low agricultural potential respectively. Soil type Av2 is the most dominant soil type, comprising approximately 30% of the study area. Soil types We1, We3 and Pn-We are shallow, greyish brown, moderately well- to somewhat poorly drained soils of which the land capability were classified as grazing potential with moderate to low agricultural potential. Soil type We1 is the second dominant soil type in the study area, comprising approximately 21%. Soil types Ar and ArR are moderately deep to very shallow, well-drained, black clay soils with shrink and expand properties of which the land capability were classified as grazing potential with moderate to low agricultural potential. Soil types My-Bo and Ms-Sw are shallow, well-drained, weak to strongly structured, clay loam soils with occasional to frequent exposed rock patches of which the land capability were classified as grazing potential with low to very low agricultural potential. Soil types Lo-Kd, Kd, Lo1 and Lo2 are yellowish grey to grey, imperfectly drained, sandy, soils underlain by soft plinthite or gleyed clay of which the land capability were classified as temporary wetland zones with low to very low agricultural potential. Soil types Se-Ka and Ka are shallow, greyish brown to grey, poorly drained, clay loam to clay soils of which the land capability were classified as temporary and seasonal/permanent wetland zones respectively, with low to very low agricultural potential. Soil types Wo, Rg2 and Rg3 are shallow, poorly drained, black clay soils with shrink and expand properties of which the land capability were classified as seasonal/permanent and seasonal/temporary wetland zones with low to very low agricultural potential. Soil type Un consists mainly of the steep edges and incised streambed of a section of the Taaibosspruit and consist of dark brown to black, clay loam, alluvial soil deposits of which the land capability was classified as a seasonal/temporary wetland zone with very low to no agricultural potential. Soil type Wb comprises a poorly rehabilitated dump which was covered with a thin layer (100-200 mm) of yellow brown, yellowish grey or grey sandy soils. The dump material appeared to be some ashy and stony discard material. Unit Q consist of quarries with no remaining topsoil which was classified wilderness land with no agricultural potential.

20 Table 2b: Soil legend for New Cornelia Vaalbank underground mining area and conveyer route SOIL LEGEND Soil Type Code Hu Bv Av1 Av2 We1 We3 Dominant Soil Form and Family Hutton 2200 Bainsvlei 2100 Avalon 2100 Avalon 2100 Westleigh 2000 Westleigh 2000, 1000 Subdominant Soil Form and Family Bainsvlei 2100, Bloemdal 2100 Hutton 2100 Bainsvlei 2200 Bloemdal 2100, 2200 Avalon 2200, Longlands 2000, Pinedene 2100 Effective Soil Depth (mm) 900-1500 1000-1500 800-1300 Longlands 2000 800-1200 Avalon 2100, Longlands 2000 Glenrosa 1121 Swartland 1111 Sepane 1110 400-800 200-400 % Clay per horizon A, E, G, B A: 22-28 B: 28-35 A: 22-28 B1: 28-35 B2: 28-35 A: 16-22 B1: 18-28 B2: 30-38 A: 5-9 B1: 6-12 B2: 28-36 A: 15-24 B: 30-38 A: 15-30 B: 30-45 Texture Class Sandy clay loam Sandy clay loam Sandy loam - sandy clay loam Sandy loamy sand Sandy loam - clay loam Sandy loam - clay Terrain Gently sloping crests and midslopes (1-2% slope) Gentle midslopes (1-2% slope) Gentle midslopes (1-2% slope) Crests and gentle midslopes (0-1% slope) Level crests or gentle midslopes (0-1% slope) Level to gently sloping (1-4% slope) crests and upper midslopes Summarized Description of Dominant Soil Form Reddish brown, sandy clay loam Orthic A- horizons, overlying sandy clay loam, well drained, red apedal or weak blocky structured B-horizons overlying weathered dolerite Reddish brown, sandy clay loam Orhtic A- horizons, overlying red, apedal or weakly structured, well-drained, clay loam B1- horizons, overlying reddish grey, mottled soft plinthic B2-horizons, overlying weathered dolerite Greyish brown, sandy loam Orhtic A- horizons, overlying sandy loam or sandy clay loam, yellow brown apedal B1- horizons, underlain by greyish, mottled, poorly drained, soft plinthic B2-horizons Yellowish grey, sandy Orthic A-horizons, overlying sandy or loamy sand, pale, yellow brown, apedal B1-horizons, underlain by greyish, mottled, poorly drained, soft plinthic B2-horizons Greyish brown, sandy loam or sandy clay loam Orthic A-horizons either directly overlying grey, mottled, poorly drained, sandy clay loam or clay loam soft plinthic B-horizons at about 400-500 mm, or via a transitional, sandy clay loam AB-horizon Dark brownish, sandy loam to sandy clay loam Orthic A-horizons, overlying via a gravel line of iron and manganese nodules, prominent mottled, fine blocky structured, clay loam to clay (pedocutanic) merging into mottled, soft, weathered shale Agricultu ral Potential Land Capability Erodibility No of Units Area (ha) Area (%) High Arable Low 4 247.74 5.139 High Arable Low 1 7.36 0.153 Moderate to high Moderate to low Low Moderate to low Arable Low 7 423.14 8.779 Arable Low 6 1459.81 Grazing Grazing Lowmoderate Low- Moderate 6 1013.19 30.28 5 21.01 9 1 6.98 0.145

21 Soil Type Code Dominant Soil Form and Family Pn-We Pinedene 2100 Ar ArR My-Bo Ms-Sw Lo-Kd Arcadia 1200 Arcadia 1200 Mayo 1100 Bonheim 1220 Swartland 1121 Mispah 1100 Longlands 1000 Subdominant Soil Form and Family Westleigh 2000, Avalon 2100 Effective Soil Depth (mm) 700-1100 % Clay per horizon A, E, G, B A: 12-20 B:13-25 Rensburg 2000 400-900 A: 45-60 Clay Arcadia 1100, Milkwood 1000 Valsrivier 1122 Swartland 1121 Arcadia 1100 Mayo 1100 Milkwood 1000 Arcadia 1100 100-600 A: 45-60 Clay 200-700 200-500 Kroonstad 1000 700-900 A: 33-40 B: 35-40 A: 30-35 B: 35-38 A: 10-20 E: 10-20 B: 28-38 Texture Class Loamy sand - sandy clay loam Clay loam Clay loam Loamy sand to sandy clay loam Terrain Gentle midslopes and upper footslopes (1-2% slope) Gently sloping (2-4% slope) midslopes, upper footslopes and isolated crests Gently to moderately sloping (2-5% slope) mainly midslopes with isolated crests and upper footslopes Gentle to moderate sloping midslopes and footslopes (3-7% slope) Moderately steep to steep midslopes and footslopes (8-15% slope) Gentle footslope (1-3% slope) SOIL LEGEND Summarized Description of Dominant Soil Form Very shallow to shallow, black, strongly structured, swelling, calcareous clay (Vertic A-horizons) overlying hard or weathered dolerite with 5-70% exposed rock at places Dark grayish brown, moderate to strongly structured, clay loam A-horizons (Melanic), directly overlying weathered dolerite, or via a thin, strong blocky structured, dark greyish brown, clay loam B-horizon (Pedocutanic); 10-40% exposed rock at places Dark grayish brown, weakly structured, clay loam Orthic A-horizons, directly overlying hard or weathered dolerite, or via a thin, moderate or strong blocky structured, dark grayish brown, clay loam B-horizon (Pedocutanic); 5-70% exposed rock at places Dark grey or grey loamy sand or sandy loam Orhtic A-horizons, over grey, imperfectly drained, loamy sand or sandy loam E-horizons, underlain at by grey, mottled, poorly drained, plinthic B or G- horizons Agricultu ral Potential Low Very low Low Land Capability Grazing Erodibility Greyish brown, loamy sand or sandy loam Orthic A-horizons, either directly overlying greyish, imperfectly drained, sandy clay loam plinthic B1-horizons, or via a loamy sand, pale yellow-brown, apedal B1- horizon. Moderately deep, black, strongly structured, swelling, calcareous clay (Vertic A-horizons) overlying weathered dolerite and shale; isolated dolerite outcrops in places Lowmoderate No of Units Area (ha) Area (%) 1 41.67 0.864 Grazing Low 2 194.81 4.041 Grazing Grazing Moderatelow Lowmoderate Lowmoderate 3 282.58 5.863 4 171.85 3.566 Very low Grazing Moderate 8 210.33 4.362 Low Wetland Moderate 1 61.68 1.280

22 Soil Type Code Lo1 Lo2 Kd Se-Ka Ka Wo Rg2 Dominant Soil Form and Family Longlands 2000 Longlands 1000 Kroonstad 1000 Sepane 1220 Katspruit 2000 Katspruit 2000 Willowbroo k 2000 Rensburg 2000 Subdominant Soil Form and Family Longlands 1000 Constantia 1100 Kroonstad 1000 Katspruit 1000 Westleigh 1000 Longlands 1000, Katspruit 1000 Willowbrook 2000 Kroonstad 1000 Longlands 1000 Willowbrook 2000 Rensburg 2000 Longlands 1000, 2000 Willowbrook 2000, Arcadia 1200 Effective Soil Depth (mm) 800-1000 800-1000 300-600 300-500 300-500 250-400 300-600 % Clay per horizon A, E, G, B A: 12-18 E: 10-18 B: 30-38 A: 6-10 E: 2-8 B: 15-30 A: 3-8 E: 1-4 G: 40-55 A: 28-35 G: 40-50 B: 30-40 A: 25-30 G: 30-40 A: 30-40 G: 40-50 A: 45-60 G: 40-60 Texture Class Loamy sand - sandy loam Sandy Sandy Clay loam to clay Sandy clay loam - clay loam Clay loam to clay Clay Terrain Gentle footslope and lower midslope (1% slope) Weakly expressed, almost level, narrow bottomlands (0-1% slope) Gentle footslopes and valley bottoms (1% slope) Gentle footslopes land valley bottoms (1-2% slope) Bottomlands or wet, level, crest areas (0-1% slope) Flat to gently sloping valley bottoms (0-1% slopes) Level (0-2% slope) valley bottoms; includes narrow drainage lines and wide flood plains SOIL LEGEND Summarized Description of Dominant Soil Form Greyish brown loamy sand Orhtic A- horizons, over grey or yellowish grey, imperfectly drained, loamy sand or sandy loam E-horizons, underlain by grey, mottled, soft plinthic B-horizons Dark grey or grey sandy Orthic A-horizons, over grey, imperfectly drained, sandy E- horizons, underlain by grey, mottled, soft plinthic B-horizons Grey, sandy Orhtic A-horizons, underlain by grey, imperfectly drained, sandy E- horizons, underlain by grey, poorly drained, clay G-horizons. Greyish brown, clay loam Orthic A- horizons, directly overlying greyish, massive or weakly structured clay (Ghorizon) or via moderate blocky structured pedocutanic B-horizons with signs of wetness within 500 mm Dark grey or grey, brown mottled, sandy clay loam Orthic A-horizons, grading into grey, calcareous, poorly drained, clay loam G-horizons Dark grey, moderately structured, clay loam, Melanic A-horizons overlying calcareous, grey, poorly drianed clay (Ghorizons) grading in to weathered shale or dolerite Deep, black, strongly structured, swelling, calcareous clay (Vertic A-horizon) overlying grey, poorly drained, gleyed, calcareous clay (G-horizon) Agricultu ral Potential Land Capability Erodibility No of Units Area (ha) Area (%) Low Wetland Low 1 81.59 1.693 Very low Wetland Lowmoderate 13 138.85 2.879 Very low Wetland Low 2 1.72 0.035 Low Very low Low Very low Wetland Wetland Wetland Wetland Lowmoderate Lowmoderate Lowmoderate Lowmoderate 1 17.22 0.357 6 266.17 5.521 1 66.73 1.384 2 11.17 0.232

23 Soil Type Code Rg3 Un Wb Dominant Soil Form and Family Rensburg 2000 Rensburg 2000 Witbank 1000 Subdominant Soil Form and Family Willowbrook 2000, Katspruit 2000, Dundee, Sepane Dundee Oakleaf Alluvium Effective Soil Depth (mm) 300-600 % Clay per horizon A, E, G, B A: 40-60 G: 40-60 Texture Class Clay loam to clay 0-500 A: 30-40 Clay loam - 100-200 A: 5-10 Sandy Q - - 0 - - Terrain Level (0-2% slope) valley bottoms; river terraces; wide dry and saturated flood plains; steep river banks with gully erosion in places Mainly riverbed; Steep, incised, rocky land along the Taaibosspruit Concave - excavated gravel pits SOIL LEGEND Summarized Description of Dominant Soil Form Deep, black, strongly structured, swelling, calcareous clay (Vertic A-horizon) overlying grey, poorly drained, gleyed, calcareous clay (G-horizon) Narrow strips of undifferentiated, dark coloured, weak to strongly structured alluvium in places Poorly rehabilitated dump; Yellow brown, yellowish grey and grey, sandy A-horizons underlain by ashy and stony discard material. Agricultu ral Potential Very low Land Capability Wetland Erodibility Low; high at riverbank No of Units Area (ha) Area (%) 1 86.66 1.798 Low-none Wetland High 1 18.57 0.385 Very low - none Wilderness Moderate to high 1 2.84 0.059 No soil - exposed dolerite rock None Wilderness High 2 7.74 0.160 TOTAL 75 4820.4 100.0

24 5.2 Other soil qualities derived from soil properties The soil qualities derived from the soil properties for each soil type are summarised in Table 3. The soil qualities were rated as high, moderate and low with classification inbetween these. Table 3: Other derived soil qualities Dry land crop Soil Type Irrigation production (Code) potential potential New Cornelia Block 1 Cv Low Low Av Low Low Lo2 Low Low Lo1 Very low Very low Kd/Ka Very low Very low New Cornelia Vaalbank Hu High High Bv High High Av1 Moderate to high Moderate to high Av2 Moderate to low Moderate to low We1 Low Low We3 Moderate to low Moderate to low Pn-We Low Low Ar Moderate-low Low ArR Very low Very low My-Bo Low Low Ms-Sw Very low Very low Lo-Kd Low Very low Lo1 Low Very low Lo2 Very low Very low Kd Very low Very low Se-Ka Low Very low Ka Very low Very low Wo Low Very low Rg2 Very low Very low Rg3 Very low Very low Un None None Wb Very low - none None Q None None 5.3 Soil chemistry The soil analytical results of 20 representative samples, collected from the A-horizon (0-200 mm) and E or B-horizon (350-700 mm) are shown in Table 4. The positions of the sampling points are shown on Figure 3a and coordinates are included in Appendix D.

25 The averages of base cations [potassium (K), calcium (Ca) magnesium (Mg) and sodium (Na)] as well as phosphorus and ph were calculated and highlighted in yellow. Table 4: Soil chemical analyses Samp Point Soil Form Hor Depth K Ca Mg Na T.Acid / Acid T.Suur saturat. Resistance mg/kg mg/kg mg/kg mg/kg cmol(+)/kg % ohm P (Bray1) mg/kg ph (H 2O) AS31 Lo2000 A1 0-250 18 187 30 0.96 0 1640 1.3 6.35 E 350-700 11 104 11 0.47 6600 0.77 6.23 AY31 Av1100 A1 0-250 24 188 25 0.19 0 5400 5.7 5.67 B1 350-700 13 187 32 3.9 4150 0.85 5.8 BE25 Av1100 A1 0-250 35 195 21 0.58 0 1410 9.2 6.39 B1 350-700 17 34 6 0.05 5220 1.6 5.7 BK29 Av1100 A1 0-250 22 124 29 0.04 0 4720 4 6.19 B1 350-700 40 52 7 0.16 6430 1.3 6.04 BM17 Lo2000 A1 0-250 28 303 45 1.1 0.06 2.9792 8830 2.2 5.53 E 250-700 12 51 6 0.37 3650 1.3 5.26 B1 1100-1500 87 469 251 7.7 7680 0.56 5.28 BO25 Av1100 A1 0-250 41 195 35 1.9 0 2080 2.7 5.6 B1 350-700 29 31 7 0.66 5890 1.1 4.89 BS11 Cv1100 A1 0-250 17 60 6 1.1 0.37 48.5397 7910 11.2 5.2 B1 350-700 8 24 2 0.59 6070 2.1 5.49 BU3 Cv1100 A1 0-250 20 98 12 1.2 0 12330 13.2 5.6 B1 250-700 8 46 5 0.27 8390 10.1 5.72 E 700-1200 5 0.76 2 0.04 14160 2.3 5.38 BW9 Cv1100 A1 0-250 19 137 17 0.47 0.04 4.3853 16560 3.4 5.54 B1 350-700 14 85 13 0.72 19480 1.9 5.64 CA21 Fw1210 A1 0-250 32 335 63 0.69 0 6020 3.1 6.03 E 350-700 22 219 53 3.4 3970 1.1 6.78 Average 24 142 31 1.2 3.7 5.7 5.3.1 Soil fertility status The averages of the base cations (K, Ca, Mg, Na) as well as phosphorus and ph (highlighted in yellow, Table 4) were compared to general fertility guidelines (Fertilizer Association of South Africa, 2003) in Table 5. Table 5: Soil fertility compared to broad fertility guidelines Guidelines (mg/kg) Average calculated Low High in Table 4 (mg/kg) Status Potassium (K) <40 >250 24 Low Calcium (Ca) <200 >3000 142 Low Magnesium (Mg) <50 >300 31 Low Sodium (Na) <50 >200 1.2 Very low, which is positive Phosphorus (P) <5 >35 3.7 Low Very acid Acid Slightly acid ph(h 2 O) neutral Slightly alkaline Alkaline <4 5-5.9 6-6.7 6.8-7.2 7.3-8 >8 5.7 Acid

26 The average cation concentration values of K, Ca and Mg are low, which indicates the highly leached status of the sandy soils. The low average Na concentration of 1.2 mg/kg is positive and indicates no accumulation of sodium and implies an absence of sodic soil conditions. The average P concentration of 3.7 mg/kg is low although somewhat higher concentration in the A-horizons occurs at sampling points BE25, BS11 and BU3 (9-13 mg/kg). The somewhat higher values in the A-horizons are probably due to fertilizers applied for former crop farming or to improved grassland and pastures. The average P status of the B-horizon is much lower than the A-horizon due to the immobility of phosphorus. The average ph value of 5.7 with a minimum of 4.89 and maximum of 6.78 indicates slightly acid to acid soil conditions.

27 5.4 Land capability and agricultural potential The land capability classification of the Northern and Southern Mining areas and the Infrastructure area were compiled on 3 separate maps, Figures 4a, 4b and 4c respectively. Each figure contains an abbreviated land capability legend. A detailed land capability legend for each of the 3 areas is provided in Tables 6a, 6b and 6c. The land capability and agricultural potential were determined from the soil types and classed according to the following categories: Arable potential o Well drained soils, o Moderately well-drained soils; Grazing potential o Well drained soils, o Moderately well-drained soils, Wetland and riparian zones o Imperfectly drained soils, o Somewhat poorly to poorly drained soils, Wilderness o Eroded areas, o Rehabilitated dumps, and o Disturbed areas. 5.4.1 Land capability of New Cornelia Block 1 opencast and infrastructure areas The location and extent of land capability classes within the proposed New Cornelia Block 1 opencast and infrastructure areas are shown on Figure 4a. The land capability and agricultural potential are summarised in Table 6a.

28

29 Table 6a shows the soil types within the New Cornelia Block 1 opencast and infrastructure areas which were grouped into each land capability class, a broad description of the soil type, the agricultural potential, the area and percentage comprised by individual soil types within each land capability class as well as the total per land capability class. Table 6a: Land capability classes and agricultural potential New Cornelia Block 1 opencast and infrastructure areas LEGEND: LAND CAPABILITY Land Capability (Code) Arable (A) Grazing (G) Temporary wetland (TW) Seasonal wetland (SW) Wilderness (WDN) Internal drainage *Soil type code Agricultural potential Broad soil description Area (ha) Well-drained - - No soils in this category 0 0 Moderately well-drained Well-drained Cv Low Imperfectly drained Imperfectly drained Somewhat poorly to poorly Disturbed, eroded or rehabilitated - - No soils in this category 0 0 Av Lo2 Lo1 Kd/Ka Low Low Low Low Area (%) Total 0 0 Soil: Very deep, brownish yellow, sandy soils 155.24 18.39 Deep, brownish yellow, sandy soils underlain by slightly 424.72 50.32 mottled soft plinthite Deep, greyish yellow, sandy soils underlain by slightly 133.22 15.78 mottled soft plinthite Total 713.18 84.49 Deep, greyish yellow to grey, sandy soils underlain by 70.90 8.40 mottled soft plinthite Shallow, grey, sandy soils underlain by gleyed clay Total 70.90 8.40 59.95 7.10 Total 59.95 7.10 - - No areas in this category 0 0 Total 0 0 *See soil map, Figure 3a Grand Total 1688.06 100.0 5.4.2 Land capability of New Cornelia Vaalbank underground mining area and conveyer route The location and extent of land capability classes within the proposed New Cornelia Vaalbank underground mining area and conveyer route are shown on Figure 4b. The land capability and agricultural potential are summarised in Table 6b.

30

31 Table 6b shows the soil types within the New Cornelia Vaalbank underground mining area and conveyer route which were grouped into each land capability class, a broad description of the soil type, the agricultural potential, the area and percentage comprised by individual soil types within each land capability class as well as the total per land capability class. Table 6b: Land capability classes and agricultural potential New Cornelia Vaalbank underground mining area and conveyer route LEGEND: LAND CAPABILITY Land Capability (Code) Arable (A) Grazing (G) Temporary wetland (T) Seasonal/ Temporary wetland (S/T) Seasonal/ Permanent wetland Wilderness (WDN) Internal drainage *Soil type code Agricultural potential Broad soil description Area (ha) Area (%) Well-drained Hu High Reddish brown to red sandy clay loam soils 247.74 5.139 Bv High Reddish brown to red sandy clay loam soils 7.36 0.153 Moderately Moderate to Greyish brown to yellow brown, sandy loam Av1 high to sandy clay loam soils well-drained 423.14 8.779 Well-drained Imperfectly drained Imperfectly drained Somewhat poorly to poorly Somewhat poorly to poorly Rehabilitated dumps Excavated quarries Av2 Ar Ar/R My-Bo Ms-Sw We1 We3 Pn-We Moderate to low Greyish yellow to yellow brown, sandy to loamy sand soils Black, strongly structured, swelling, clay Moderate-low soils Black, strongly structured, swelling, clay Very low soils; 5-70% exposed surface stone Dark grayish brown, moderate to strongly Low structured, clay loam soils; 10-40% exposed surface stone Dark grayish brown, weak to strongly Very low structured, clay loam soils; 5-70% exposed surface stone Greyish brown to grey, mottled, sandy loam Low or sandy clay loam soils Moderate to Greyish brown to grey, mottled, sandy loam Low or sandy clay loam soils Yellow to greyish brown, sandy loam or Low sandy clay loam soils 1459.81 30.285 Total 2138.05 44.356 194.81 4.041 282.58 5.863 171.85 3.566 210.33 4.362 1013.19 21.019 6.98 0.145 41.67 0.864 Total 6197.51 128.57 2 Dark grey or grey loamy sand or sandy Lo-Kd Low loam spoils 61.68 1.280 Greyish brown to yellowish grey loamy Lo1 Low sand soils 81.59 1.693 Lo2 Very low Dark grey or grey sandy soils 138.85 2.879 Greyish brown to grey, structured clay loam Se-Ka Low soils 17.22 0.357 Rg2 Rg3 Un Very low Very low Low-none Deep, black, strongly structured, swelling clay soils Deep, black, strongly structured, swelling clay soils Incised drainage channel with black strongly structure clay soils Total 12694.36 263.35 3 11.17 0.232 86.66 1.798 18.57 0.385 Total 25505.12 529.12 1 Kd Very low Grey sandy soils 1.72 0.035 Ka Very low Grey sandy soils 266.17 5.521 Wo Low Brown to black, structured clay soils 66.73 1.384 Total 51344.86 1065.1 82 Wb Very low - none Greyish yellow to grey sandy soils manmade soil horizons 2.84 0.059 Q None No topsoil 7.74 0.160 Total 102700.3 2130.5

32 *See soil map, Figure 3a Grand Total 4820.4 100.0 83

33 5.5 Wetland and riparian delineation Land capability was assessed in categories of arable land, grazing land, wetland and riparian zones and wilderness land. Wetland and riparian zones were therefore delineated as part of the land capability assessment based on soil properties by means of systematic auger observations towards wetlands and riparian zones in order to locate the point where soil properties reflect signs of wetness within 500 mm from the surface or where soil, topography and vegetation combined, indicate the boundary of the riparian zone. The soil types associated with wetlands and riparian zones are summarized in Tables 6a and 6b and the locality and extents and are shown on the land capability maps Figures 4a and 4b. A 100 m wetland buffer zone is indicated with a blue line. (See Appendix C for details on soil properties related to wetland zones). 5.6 Pre-mining land use 5.6.1 Pre-mining land use within the New Cornelia Block 1 opencast and infrastructure areas The locality and extent of current land uses within the New Cornelia Block 1 opencast and infrastructure areas are shown in Figure 5a and are summarized in Table 7a.

34

35 Table 7a: Pre-mining land uses within the New Cornelia Block 1 opencast and infrastructure areas Land Use Pre-mining Land Use Area (ha) Area (%) Code G Grazing Commercial cattle farming 813.57 96.39 SM Sand mining (approximate locality and extent) 30.01 3.56 TR Tar road 0.44 0.05 TOTAL 844.02 100.0 5.6.2 Pre-mining land use within the New Cornelia Vaalbank underground mining area and conveyer route The locality and extent of current land uses within the New Cornelia Vaalbank underground mining area and conveyer route are shown in Figure 5b and are summarized in Table 7b. Table 7ba: Pre-mining land uses within the New Cornelia Vaalbank underground mining area and conveyer route Land Use Pre-mining Unit Area (ha) Area (%) Code Land Use Count M Maize 9 1149.55 23.85 SB Soybeans 3 601.80 12.48 S Sorghum 1 8.34 0.17 P Pasture 8 482.30 9.99 G Grazing 7 2279.88 47.30 FL Fallow land 4 150.44 3.12 T Trees 5 41.14 0.85 FS Farmstead 9 41.39 0.85 H Housing 3 7.50 0.15 R Road 4 22.59 0.47 V Vacant 6 35.21 0.73 RD Residential 1 0.25 0.01 TOTAL 60 4820.39 100.0

36

37 5.7 Historical agricultural production Crop yields vary from farm to farm and field to field due to various factors such as soil types and climatic conditions e.g. annual precipitation, temperature, day lengths, heat units etc. No crop farming occurs at New Cornelia Block 1 and the total area is utilized for grazing purposes. Some crop farming could have taken place at some stage although the crop yields would be fairly low due to the very sandy soil texture and subsequent low water holding capacity. Carrying capacity was broadly assessed based on dominant grass species and basal cover during the time of the soil assessment and is summarized in Table 8a. Table 8a: Historical agricultural production New Cornelia Block 1 Derived soil Product Soil Types) potential Grazing - Cattle All soil types (including wetland) Potential Yield (ha/lsu) Low 5-7 Crop farming within the New Cornelia Vaalbank study area occurs mainly on the red sandy clay loam soils and brownish yellow to greyish yellow sandy to loamy sand soils. The brown and black clay soils and grey sandy wetland sections are utilized for grazing purposes. Potential crop yields were estimated based on soil properties and summarised in Table 8b. Table 8b: Historical agricultural production New Cornelia Vaalbank Derived soil Potential Yield Product *Soil Types potential (tons/ha/annum) Hu, Bv, Av1 Moderate to high 5-8 Maize Av2, We1 Moderate to low 2-4 Soybeans Hu, Bv, Av1 Moderate to high 1.5-2 Av2, We1 Moderate to low 1-1.5 Grazing - Cattle We3, Pn-We, Ar, ArR, My-Bo, Ms-Sw, Lo-Kd, Lo1, Lo2, Se-Ka, Kd, Ka, Wo, Rg2, Rg3 Low-very low 4-7 ha/lsu * See soil map Figure 3b 5.8 Evidence of misuse No evidence of misuse was observed.

38 6. PROCEDURES FOR STRIPPING AND HANDLING OF SOIL 6.1 Principles in stripping and handling of topsoil The term topsoil refers to the A and B-horizons of the soil profile as defined in the Taxonomic Soil Classification system for South Africa. The A-horizon comprises the upper part (0-300 mm) of the soil profile and the B-horizon from 300 mm up to the stripping depth specified per soil type indicated in Figure 6. The topsoil and A- and B- horizons are further described in Appendix E in terms of soil stripping, stockpiling and replacing. Stripping, stockpiling and replacing of topsoil has a very high impact on soil, land capability and land use and the procedures followed during execution of these actions directly influences the post-mining land capability and consequently determines the degree of deterioration from pre-mining to post-mining land capability. It also directly determines the possible post-mining land uses. Stripping and stockpiling of topsoil, shaping of spoils and replacing of topsoil should therefore take place according to a plan which should be well managed and progressively adapted according to circumstances. The following should be included in topsoil stripping and stockpiling plans: Prevent stripping and mixing of high quality topsoil (A and B-horizons) with low quality underlying material to ensure sufficient volumes of high quality soil for rehabilitation. The quality of soil earmarked for rehabilitation purposes significantly deteriorates when the high quality topsoil is mixed with the underlying poorer quality material (clay layers, calcrete, plinthite, weathered rock etc). This results in significant deterioration in the quality of soil physical and chemical properties and a decline in soil fertility necessary for revegetation. The deterioration in soil quality also significantly increases the susceptibility of rehabilitated soils for erosion and seal and crust formation. Separate stockpiling of different soil types to obtain the highest postmining land capability. Topsoil quality or potential is not just limited to the grade of soil generally referred to as topsoil but can vary from very high to low due to various properties. Soil properties of different soil types can vary substantial e.g. high quality red and yellow well-drained soils and low quality grey poorly drained wetland soils over very short distances in the same field. Mixing of different soil types result in rapid changes in soil properties and characteristics such as texture, infiltration rates and water holding capacity over short distances after replacement, which will definitely adversely affect the post-mining land capability. Separate stripping, stockpiling and replacing of soil horizons (A and B- horizon) in the original natural sequence to combat hardsetting and compaction, maintain soil fertility and conserve the natural seed source. The higher soil fertility of the A-horizon, especially phosphorus and carbon contents, declines significantly when it is mixed with the B-horizon resulting poorer re-vegetation success. It also increases the susceptibility to compaction and hard setting. The A-horizons also serves as a seed source which will enhance the re-establishing of natural species. The A and B- horizons should be stripped and stockpiled separately and replaced with the A-horizon overlying the B-horizon.

39 Contrary to the general perception, separate stockpiling of different soil types and horizons does not have significant cost implications for the mine and only requires planning and continued management. It is therefore crucial to strip according to guidelines and depths as described in the following sections. The soil horizons and properties influencing stripping and stockpiling procedures are discussed in Appendix E. 6.2 Construction phase The erection of mining infrastructure will have impacts on the soil, land capability and land use during the construction phase. The following procedures for topsoil stripping and handling during the operational phase should be followed: Buildings and conveyers. All excess topsoil which might be excavated for the foundations of these structures should be stored for later rehabilitation. Sidings, haul roads and coal stockpiles. Soil pollution should be prevented at all times. Wherever coalliferous material or any material with a potential polluting ability will be dumped and handled, the footprint should be covered with at least 300 mm subsoil or soft overburden material and the edges should be elevated (berm) to prevent pollution beyond the footprint. Discard, tailings and waste dumps. At least 300 mm of topsoil should be removed and stored for later rehabilitation. Borrow pits may not be made on a later stage in order to get topsoil for rehabilitation of the dump. The footprint should be compacted and sealed/lined, it should be free draining and all seepage should be channeled to pollution control dams. Pollution control dams. The A-horizon (0-300 mm) and B-horizon of the topsoil should be stripped and stored separately for later rehabilitation. Coalliferous materials with a potential polluting ability may not be used to stabilize the base or wall embankments of the dam. The base of the dams should be lined. 6.3 Operational phase (Open Pit) In practice, even with optimal rehabilitation procedures applied, deterioration from premining to post-mining land capability is unavoidable. It is therefore crucial to follow the proposed rehabilitation procedures precisely in order to prevent degradation of soil characteristics and to re-establish the highest possible post-mining land capability. The following stripping, stockpiling and replacing procedures needs to be executed precisely: Figure 6 and Table 9 show the soil types in the proposed New Cornelia opencast area that needs to be stripped and stockpiled on 2 separate stockpiles. The boundaries of the soil types that should be stripped at different depths and/or stockpiled separately, as shown in Figures 6, should be surveyed and staked at 50 m intervals before any soil stripping commences. This can be done by the surveyor or soil specialist. The stake points are shown in Figure 6 as yellow dots at intervals of 50 m and the coordinates are provided in Appendix D. The stripping plan, Figure 6 includes small wetland zones. Mining of these areas are subject to authorization of the relevant government departments. The proposed topsoil stockpiles positions, consisting each of a section for the A and B-horizon needs to be surveyed and staked by the mine surveyor. The most suitable stockpile positions should be determined by the mining engineer considering the mining block sequence of the mine plan.

The A and B-horizon should be stripped and stockpiled separately as specified in the guidelines for the rehabilitation of mined land (Chamber of Mines of South Africa, 2007). The A and B-horizon sections should be marked with a signboard. The size of the stockpiles should be based on the soil volume per stockpile as indicated in Figures 6 and no height restriction is proposed. Before topsoil can be replaced, the pit should be backfilled to an elevation similar to the pre-mining topography as far as possible with the material available, in order to ensure a continuation of the pre-mining surface drainage pattern. The backfilled surface should be surveyed by a surveyor in order to ensure that it has the correct elevation and slopes, and is free draining. A non free draining surface results in local depressions of periodically saturated zones and increased percolation which usually leads to localised subsidence of underlying spoil material. Slopes of the spoil surface should therefore aim to be similar to the pre-mining surface and should change gradually since abrupt changes in slope gradient increase the susceptibility for erosion initiation. The backfilled surface should first be replaced with soil of the B-horizons. Care should be taken to tip enough soil per square unit to reinstate the total required B-horizon thickness at once. The dumped soil heaps should thus only be levelled to reach the required horizon/soil thickness. This will prevent compaction in the lower profile which cannot be alleviated afterwards. Spreading of soil over far distances and repeated traversing should be avoided. Caterpillar tracked equipment is preferred to use for levelling of the B-horizons due to the large quantities of soil that needs to be handled and less compaction caused by these tracks. Bowl scrapers cause enormous compaction and may not be used. When the roll-over mining system kicks in and the point is reached where soils are stripped and directly replaced, without stockpiling, the following method should be implemented. The A-horizon of 1 mining strip should be stored at the final mining strip and the B-horizon should be tipped and levelled on the area to be rehabilitated (properly levelled spoil surface). The A-horizon of the next mining strip should then be tipped and levelled on top of the replaced B-horizon and the roll-over system can continue like this. The stored A-horizon of the first mining strip should then be replaced on the B-horizon of the last mining strip. After the B-horizon is replaced, the surface should be loosened to a depth of approximately 300 mm with normal agricultural equipment, preferably a multiple teeth implement. This is very important to prevent a compacted layer between the A and B-horizons which will be similar to a plough sole which dramatically reduces the effective soil depth and restrict root development. The A-horizon should then be tipped systematically over the loosened B- horizons surface and spread evenly. Replacing the A-horizon involves much smaller quantities of soil and a combination of a lighter dozer and grader should be used. Graders have the ability to create a more even surface with less traversing than a dozer, without creating too much compaction. The replaced topsoil thickness should progressively be monitored during replacement to verify if it is similar to the replacing depth provided in Figure 6a and 6b and to prevent that shortages of topsoil are not encountered. The soil fertility status should be determined by soil chemical analysis after levelling (before seeding/re-vegetation), and soil amelioration should be done accordingly as recommended by a soil specialist, in order to correct the ph and nutrition status before re-vegetation. The rehabilitated sections should be re-vegetated with a grass mixture dominated by local climax species in early summer to stabilize the soil. A short term fertilizer program should be based on the soil chemical status after the first year in order to maintain the fertility status for 2 to 3 years after 40

41 rehabilitation until the area can be declared as self-sustaining. The soil types that should be stripped and stockpiled either separately or together are shown in Figure 6. The figure includes a table that shows the stripping depths, the areas and percentages as well as the total soil volume per soil type based on the stripping depth. It also shows the replacing depth (topsoil thickness) and post-mining land capability class. This was determined by calculating the total soil volume per soil group (stockpile), divided by the original area which was stripped. This implies that if more than 1 soil type, which were stripped at different depths, are stockpiled together, it will be replaced at 1 average depth. It further shows the stake points on soil boundaries at 50 m intervals. The criteria for soil stripping and replacing within the New Cornelia opencast area are summarised in Table 9 for the various soil types. Table 9: Criteria for soil stripping and replacing at New Cornelia Block 1 opencast Legend: Soil stripping depths, volumes, post-mining replacing depths and land capability Soil group and stockpile no. Stockpile 1 (Brownish and greyish yellow, moderately well-drained soils) *Stockpile 2 (Grey, imperfectly to poorly drained wetland soils) Soil Types Strip depth (m) Pre-mining Area (ha) Area (%) Soil volume (m 3 ) Cv 1.4 139.70 21.56 1955800 Av 1.0 263.79 40.71 2637900 Lo2 1.0 129.63 20.01 1296300 TOTAL 533.12 82.28 5890000 Lo1 0.3 55.49 8.56 166470 Post-mining Replacing Depth (m) Area (ha) Land capability 1.1 533.12 Grazing Kd/Ka 0.3 59.36 9.16 178080 0.3 114.85 Wetland/ Grazing TOTAL 114.85 17.72 344550 Grand Total 647.97 100.00 6234550 647.97 *Stripping of the wetland soil units is subject to authorization by the relevant Government authorities

42 6.3.1 Post-mining land capability commitments The post-mining land capability class will mainly be determined by the soil type and the thickness of the soil layer placed back on the spoil surface. Other factors and characteristics that might influence the post-mining land capability is slope, compaction and reduction of soil quality due to contamination of soils by subsoil, soft overburden or spoil material. A post-mining land capability assessment needs to be done progressively (annually) during the operational phase by a soil specialist by means of auger observations at a