An efficient production is essential for an economic continuation of most gold mining companies, due to a moderate gold price and a competitive market. Recently, the Ahafo South Mining operations owned by the Newmont Mining Corporation and located on a Proterozoic gold belt in the south west of Ghana, indicated that the cost of production must be reduced in order to viably continue the mining operations. For this research, the Ahafo South Mine Site was used as a case study. A site visit of 90 days provided insight in where the efficiency of productivity could possibly be improved. During this visit, historical production- and financial data [2014-2016] were gathered from a diverse group of engineers on site. Several audits were conducted to form a solid hypothesis with regards to the optimization of rock fragmentation. The hypothesis of this research states that improved rock fragmentation will result in a reduction of operational costs and delays along with an increase in crusher- and mill throughput.

This case study was used to investigate the effect of a varying fragmentation on shovel loading time and crusher- and mill performance. The financial impact of a potential blast pattern change was also investigated. After a statistical analysis of the available data on post-blast fragmentation, it was observed that the desired run of mine (ROM) P80 target of 350 mm is rarely met. It was concluded that the "Kuz-Ram" model used to predict the fragment size distribution is incorrectly applied, which resulted in a planned blast design that produces a coarse fragmentation with an excessive presence of boulders. This is referred to as "bad fragmentation" and proved to contribute significantly to redundant wear-and-tear of mining equipment, delays at the shovel loading operations and the primary crushing station. A correlation between fragmentation and loading time was found after a statistical analysis on the daily average loading time of three Liebherr 9400 shovels. It was concluded that the shovels are only able to produce at the desired rate of 2.4 minutes per truck load when the fragmentation of the ROM meets the P80 target. The Six Sigma principle Define, Measure, Analyze, Improve, Control (DMAIC) was applied to define an accepted interval range of shovel loading time variation (2.4 m. ± 20 s.). This so called target interval was generally not achieved by the shovels. All three shovels show an increase in average loading time during period where the ROM fragmentation was coarser.

Data generated by a VisioRock™ camera showed, in combination with the throughput of the mills, that an appropriate fragmentation is crucial for an efficient milling process. A detailed analysis of the available data, categorized in "periods of high interest", demonstrated that the power consumption of the mills significantly increased when the feed consisted of relatively coarse fragments. It was observed that the post-crushed fragmentation was more coarse when a high fraction of the feed originated directly from the pit. This observation proved that the primary crusher is not able to eliminate all boulder related issues prior to the crushing operations as a result of inadequately planned blasts. Due to inaccurate or unreliable data on mill feed, it was concluded that a full mine to mill survey is required to obtain data needed for an accurate optimization of mill performance.

Newmont Ghana Gold Limited (NGGL) stated that at least 2.3M US$ has been saved on a yearly basis since 2015, as a result of the reduction of explosives used through a blast pattern widening. The cost impact analysis showed that it is uneconomical to save costs on explosives by widening the blast pattern, due to the potential downstream effects when the fragment size distribution is affected. The financial data was inconsistent and it is therefore recommended to conduct a detailed analysis to provide insight in the actual overall (financial) benefits of a blast pattern optimization. It is recommended to investigate both "double-benching" and the purchase of a continuous surveying camera system. Another recommendation is to narrow the blast pattern down to 3.5 m. by 4 m. so that fragmentation related issues are avoided which are currently to be expected for the continuing production in the Subika- and Awonsu Pit.

Up until now, global mining companies have failed to create a fatality free environment for their employees, whether it be through increasingly difficult conditions or inadequate safety management approaches. Desiring to maintain the social license to operate, these companies are now looking into alternate ways of safety management with the aim of mitigating fatalities in the workplace.
In this study, a new risk management approach was applied to mitigate fatal incidents through the utilization of critical controls. The aim of this study was to create a scalable, minimally invasive proof-of-concept for AngloGold Ashanti that can successfully be implemented at any of the company’smining operations.
The foundation for this implementation was laid through a literature study blending contemporary safety and risk management approaches with Business Intelligence systems thinking and a perspective on industry best-practices. Subsequently, a framework of organizational requirements was set up based on 26 interviews with stakeholders in various departments within the company. Both design and implementation were conducted on-site at an AngloGold Ashanti operation in Western Australia.

The system was designed by adhering to organizational requirements, and ensuring that it is suitable to any mining environment. The designed Critical Control Management System was subsequently implemented at Sunrise Dam, one of AngloGold Ashanti’s Australian mining operations. To ensure that critical controls were also assessed at the operational level, a workplace inspection process was modified to generate control data. All sources of data subsequently were fed into a Business Intelligence environment enabling insight into critical control performance to all company stakeholders. Doing so informs decision-making on safety priorities company-wide, based on real-time data generated on the operational level.

Two case studies were performed to assess two of the most significant hazards at Sunrise Dam. The studies showed that the effectiveness of reactive controls changes irrespective of their compliance and performance. Furthermore, the influence of human factors within risk management remains difficult to quantify. Finally, it demonstrates the potential for integration of incident data into the Critical Control Management System, thus creating both leading and lagging indicators for safety performance.

The conclusion of this study is that an effective and scalable Critical Control Management System can be successfully implemented in a mining operation if the right conditions are generated. The approach of integration in existing processes demonstrates that companies can achieve greater control over fatality prevention without the need for an additional safety management system. On this basis, it is recommended that other operations are supported in creating an environment suitable for adaptation before Critical Control Management is implemented.