APPENDIX
Independent assurance statement

To the board and stakeholders of Impala Platinum Limited

Integrated Reporting & Assurance Services (IRAS) was commissioned by Impala Platinum Limited (hereafter, ‘Implats') to provide independent third party assurance (ITPA) over the information contained within this Sustainable Development Report (hereafter, ‘the SDR'), covering the 12-month period from 1 July 2013 to 30 June 2014. The assurance team comprised primarily of Michael H Rea, our Lead Certified Sustainability Assurance Practitioner (CSAP), with more than 15 years' experience in environmental and social performance measurement and assurance in over 75 assurance engagements in 17 countries, Jordan Pruner (Associate CSAP), our team of junior associates, and additional technical support from Petrus Gildenhuys (Gilden Assurance), Dr Markus Reichardt: both of whom have more than 10 years' experience in sustainability reporting and assurance.

AccountAbility AA1000AS (revised, 2008)

To the best of our ability, this assurance engagement has been managed in accordance with AccountAbility's AA1000AS (2008) assurance standard, where the format of the engagement was structured to meet the AA1000AS Type II (Moderate) requirements.

Independence

IRAS was not responsible for the preparation of any part of this report and has not undertaken any other engagements for Implats that would compromise our ability to afford ITPA over this year's report. Responsibility for producing this report belonged to Implats and its sustainability reporting advisers (Incite). Thus IRAS is, and remains, an independent assurer over the content and processes pertaining to this report.

Although this statement is directed at key Implats stakeholders, as the users of the report, IRAS' ultimate responsibility in performing its assurance activities is to the management of Implats in accordance with the terms of reference agreed with them.

Assurance objectives

The assurance process was designed to provide stakeholders of Implats with an independent ‘moderate level assurance' opinion on whether the report meets the following objectives:

Adherence to AccountAbility's AA1000AS (2008) principles of Inclusivity, Materiality and Responsiveness
Alignment with the Global Reporting Initiative (GRI) G4 guidelines, with the objective of establishing whether or not the report has met the core level of reporting requirements
Reasonable adherence to local and international expectations for effective reporting, including guidance provided through the King Code of Corporate Governance (King III) and the International Integrated Reporting Committee (IIRC) recommendations for integrated reporting
Accuracy, consistency, completeness and reliability of reported data, relative to a sample set of selected sustainability indicators, as tested at selected sites
Completeness and reasonability of selected sustainability indicators contained within the report, relative to the 122 data points reviewed for IRAS's Sustainability Data Transparency Index (SDTI).

In meeting the Type II assurance objectives, IRAS undertook a review of selected sustainability performance indicators at the following operational sites: Rustenburg Operations, Zimplats (Mine and Smelter), Springs Refinery and Mimosa. A desktop review of Mimosa's safety data was also conducted while at Zimplats' operations.

Scope of work performed

AA1000AS (2008) Compliance

The process used in arriving at this assurance statement is based on AccountAbility's AA1000AS (2008) guidance, as well as other best practices in assurance. Our approach to assurance included the following:

A review of sustainability measurement and reporting procedures at Implats' head offices and Rustenburg Operations to determine the context and content of sustainability management by the company
A review of Implats' information collation and reporting procedures to define the content of the report by looking at the materiality of issues included in the report, stakeholder engagement responses to issues identified, determination of sustainability context and coverage of material issues, ultimately leading to adherence to the AA1000AS principles of Inclusivity, Materiality and Responsiveness
A review of data collection, collation and reporting procedures at selected sites (as stated above), with specific reference to the following selected sustainability performance indicators:

Indicator		Quantity	Page(s)
1. PHW	Person Hours Worked: The number of hours worked – calculated or counted via clock-in systems – for all employees and contractors	94 056 773	101
2. LTIFR	Lost Time Injury Frequency Rate: The calculated number of employee and/or contractor injuries on duty resulting in an ability to return to work the next calendar day per 1 000 000 PHW	3.92	5, 55, 101
3. Fatals	Fatal Injuries: The number of employee and/or contractor injuries on duty resulting in death	4	4, 53, 100
4. TIFR	Total Injury Frequency Rate: The calculated number of employee and/or contractor injuries on duty resulting in either an ability to return to work the next calendar day, the need to obtain medical treatment beyond basic first aid, or fatalities, per 1 000 000 PHW	11.90	5, 55, 101
5. NIHL	Noise Induced Hearing Loss: The number of new cases of hearing loss greater that a loss of hearing (or ‘PLH shift') greater than 10%, deemed to have been the result of employment activities and submitted for consideration for compensation	36	58, 102
6. TB	Tuberculosis: The number of new cases of Tuberculosis, regardless of whether or not the case is deemed “occupation-related”	268	58, 102
7. VCT	Voluntary Counselling & Testing: The number of tests conducted of employees and contractors for HIV/Aids as a result of counselling provided by medical personnel or external specialists	10 086	58, 102
8. ART	Antiretroviral Treatments: The number of employees who are enrolled in ART programmes at year-end, net of any enrolments that were terminated – for all reasons – during the year	4 276	58, 102
9. Electricity	Total direct and indirect electricity consumption – from all sources – for primary purposes (i.e., mining and processing activities)	2 780 000 MWh	91
10 . Petrol	Total direct and indirect petrol consumption, inclusive of vehicles, generators and other equipment (i.e., for all uses)	830 000 Litres	104
11. Diesel	Total direct and indirect diesel consumption, inclusive of vehicles, generators and other equipment (i.e., for all uses)	21 776 000 Litres	104
12. Coal	Total direct and indirect coal consumption, for all purposes	127 883 Tonnes	104
13. Total energy consumed	Total direct and indirect energy consumed from all sources	14 395 000 GJ	91, 104
14. Water	Total direct and indirect water consumption – from all sources – for primary purposes (i.e., mining and processing activities)	34 775 Mℓ	88, 89, 104
15. Indirect CO₂	Total indirect Carbon Dioxide emissions	2 714 157 Tonnes	91, 105
16. Indirect NO_x	Total indirect Nitrogen Oxide(s) emissions	12 107 Tonnes	92
17. Indirect SO_x	Total indirect Sulphur Oxide(s) emissions	24 665 Tonnes	92
18. Direct CO₂	Total direct Carbon Dioxide emissions	323 997 Tonnes	91, 105
19. Direct SO_x	Total direct Sulphur Oxide(s) emissions, at Rustenburg Operations, Zimplats and Springs Refinery	30 735 Tonnes	92, 105
20. Procurement	Presence and application of systems and controls for measuring and managing procurement from ‘historically disadvantaged South Africans', as per the Employment Equity Act of South Africa, inclusive of rand value and percentage of total spend	R2 015 000	82
21. SED spend	Total rand value of expenditures on socio-economic development (SED) projects, inclusive of expenditures against Social & Labour Plan commitments	R71 million	74
22. HDSA in management	The number of persons deemed ‘Historically Disadvantaged South Africans', as per the Employment Equity Act of South Africa, working in management positions, above the supervisor level – excluding non-executive directors	378	66
23. Females in management	The number of women working in management positions, above the supervisor level – excluding non-executive directors – in South Africa	146	66

Reviews of drafts of the report for significant errors, anomalies and/or insupportable assertions
A series of interviews with the individual responsible for collating and writing various parts of the report in order to ensure selected claims were reported and substantiated.

Due to the scope and nature of this AA1000AS (Type II, Moderate) assurance engagement, the site visits were undertaken to test the authenticity of data at the primary source of collection and collation, and the report – inclusive of the sustainability information contained within both the SDR and IAR – has been assessed at the point of data aggregation for accuracy of reporting.

GRI Compliance

In determining the GRI G3 ‘application level' of the report, we performed the following exercises:

A review of the process used to define the content of the report by looking at materiality of issues included in the report, stakeholder engagement response to stakeholder issues identified, determination of sustainability context and coverage of material issues.
A review of the approach of management to addressing topics discussed in the report.
Reviews of the report to confirm that the requisite number of Implats has met the GRI's G4 ‘Core' application level requirements, inclusive of the need for reasonable discourses on materiality and disclosures on management approach to data measurement, monitoring and reporting.

Findings

Overall, it was determined during the assurance process that, in general, the company's sustainability reporting processes are comprehensive, and it was noted that:

Implats has developed comprehensive systems for data collection, collation and reporting, at both the operation and Group level, and appear to be in a constant state of continuous improvement through the development and implementation of enhanced sustainability data policies, procedures, systems and controls. While the ongoing reliance on multiple Excel spreadsheets to collect, collate and report data continues to create opportunities for simple data transcription errors – affecting the accuracy and/or reliability of some data at some collection points at the sites – procedures appear to be improving to address these potential problems.
Aside from the following exceptions, the tested site-specific data was found to be reasonably accurate and/or reliable, although process improvements at some sites may still be required with respect to the implementation of internal control procedures for data accuracy and reliability. Exceptions:
- The way in which Medical Treatment Cases (MTCs) are recorded differs among the sites, with one of the five tested sites recording MTCs inclusive of First Aid Cases (FACs), and the others excluding FACs from the total number of MTCs. This results in non-comparable data throughout the Group, and a tendency towards over-reporting the total number of MTCs, thus making TIFR data non-comparable to peer companies.
- Implats' applied definition for Lost Time Injuries (LTIs) is exclusive of Restricted Work Cases (RWCs), which is inconsistent with standard local and international practice within the Metals & Mining Sector, thereby resulting in data that is not comparable to peer injury frequency rates (i.e., LTIFR).
- Voluntary Counselling and Testing (VCT) rates for HIV/Aids – based on current data collation systems and controls – must be deemed not comparable beyond the limits of historical comparisons of Implats' prior year data, due to instances of multiple counting of tests conducted, rather than the number of persons tested.
- Due to changes in the way in which the sulphur content within concentrate is assessed at Zimplats, there has been a significant increase in total SO₂ emissions at that site. As such, current data should be deemed ‘improved', yet not wholly comparable to prior year data.
Implats adequately adheres to the AccountAbility AA1000APS principles of Inclusivity, Materiality and Responsiveness.
The report adequately meets the GRI's G4 requirements for Core level reporting.
As per IRAS's own Sustainability Data Transparency Index (SDTI), Implats scores extremely well compared to most JSE-listed companies, demonstrating an active commitment to data transparency relative to reasonably expected key data points.

Conclusions and Recommendations

Based on the information reviewed via our site visits, desk research and management interviews, IRAS is confident that the sustainability information reported within the SDR and IAR provides a comprehensive and balanced account of Implats' environmental, safety and social performance for the period under review. The data presented is based on a systematic process and we are satisfied that the reported performance data accurately represents Implats' ability to manage and/or report on its performance, while meeting the AA1000AS (2008) principles of Inclusivity, Materiality and Responsiveness. Moreover, and although the quality or quantity of data of some GRI G4 indicators can be improved, this report appears to meet the GRI's G4 Core level requirements.

However, the following recommendations have been identified:

With respect to the issue of non-comparability of safety data to industry peers, Implats should continue to investigate the need for updating its currently applied definitions, and update those definitions where deemed necessary.
Implats should continue to ensure that future assurance engagements include the site visit testing of data, vis-à-vis the expectations of AA1000AS (Type II), but because of the long history of assurance over sustainability data, should consider reducing the frequency of site visits to cover all sites in a two or three year cycle, relying on desk review in interim years. The excellent systems and controls currently in place – compared to most other companies – suggests that sites can demonstrate that they adequately meet the tests for accuracy, consistency, completeness and reliability, and therefore do not need annual assurance site visits.

For more information about the assurance process employed to assess the sustainability information contained within Implats' 2014 Sustainable Development Report, email michael@iras.co.za.

Integrated Reporting & Assurance Services

Johannesburg

11 September 2014

APPENDIX Independent assurance statement