Ya Rona Mine Accident: briefing by Goldfields

Meeting report

Opening remarks
The Chairperson said the Committee would receive a briefing on the fatalities at Ya Rona where five people had died the previous month. Members had only heard about this accident in the news, and this was an opportunity for them to get the information first hand. An investigation into the accident was underway. The briefing needed to update them on what was happening, as reports indicated that the fire was still raging underground. The delegation was at liberty to clarify the information given out via the media. Details about the accident would assist the Committee in making a determination on whether to visit the mine.

He outlined the Committee's programme for the term, and it involved an oversight to KwaZulu Natal and a trip to Australia. The Committee had been delaying the trip to KwaZulu Natal where it was expected to conduct oversight on progress on the rehabilitation of mines, and if mines complied with the law. In terms of the programme Members would have to double their efforts as the programme looked detailed.

Mr J Lorimer (DA) asked if the safety aspect would be included when the Committee listened to stakeholders in August brief them on the dip in platinum prices and the closure of mines.

The Chairperson replied there was no need as the topic was the dip in platinum prices and the closure of mines. The issue of safety could be dealt with at the briefings if the stakeholders brought it up, but the briefings were only meant to look at the international dip in the price of platinum.

Ya Rona accident
Mr Mark Roebert, Acting Senior Vice President, Goldfields Kloof Driefontein Complex (KDC), said an underground fire had occurred, at about 2855m deep, on 30 June 2012. Operations were as per normal, and emergency teams were dispatched to rescue the workers. Most of the nightshift staff were recovered successfully, but five of them (Mlungiselwa Madevana; Colekele Bankane; Musa Dlamini; Mziwakhe Mankunzi;Bavuyise Mbola) were found dead.

The normal operations started at about 17h00 on 30 June. Fire was detected at around 18h45 from shaft four. The central control fire detection heads (smoke monitors), connected to all fire detection systems in all shafts, set off an alarm. As per normal, following the three-hour grace period allowed after blasting, tramming crews (drivers of locomotives and cleaners of broken rock out of the rails) proceeded underground to do mud loading on 39 level. This was an overtime shift. It was a common occurrence to have Saturday nightshifts, and was permissible by law and was sanctioned by the Department of Mineral Resources (DMR). Specific work would be done on maintenance of underground railways, tunnels and tramming ways and this did not involve actual mining.

Reports of people collapsing on 39th level, about 2.5 km from the shaft, were received at about 23h36. Instructions were given by two senior officials - a ventilation officer and an underground manager - regarding beacon heads and critical fans. Teams were called out. The response time by the rescue teams was the normal hour and half.

Mr Roebert said 15 people were rescued at about 03h00 and were taken out of the mine. The rescue team established a fresh air base, and proceeded to search for the missing persons. At about 05h00, the five missing employees were located and they had already died. Gas emission readings indicated that Carbon Monoxide (CO) stood at 25000 parts per million (ppm), and Carbon Dioxide (CO2) was 55000 (ppm).

DMR was notified of the fatalities, and its inspectors subsequently arrived at the mine. At about four in the afternoon, the bodies of the deceased were successfully retrieved and brought to the surface. Some of the facts established were that the deceased were found in close proximity to a refuge bay (place of safety). He explained that refuge bays were a legal requirement and were established at a distance of 500m on all sites of operations. This was life sustaining and provided access to basic amenities like water, fresh air, and a telephone.

All the deceased were also found with self-contained self rescuer equipment, but none had used the self rescuer kits. The condition at the mine was that every person going underground had to don a self-rescuer kit for an eventuality of noxious gases. This kit allowed a person fresh air for a period of 30 minutes whilst searching or trying to access the place of safety. Since 2001, everyone at KDC operations had been issued with a self-rescuer kit.

The four of the deceased had gas detection instruments (GDIs) at the time of the accident. The purpose of the GDIs was to detect increased levels of life-threatening gases, and provide early warning. When the reading on this piece of equipment reached 50, it would set off an alarm, and one had to find a refuge chamber.

Remedial action
Mr Roebert said management had immediately stopped operations when it heard of the incident. Operations resumed on 3 July at KDC East, after consultation with DMR, and on 9 July at KDC West. Mining activity at Ya Rona Mine remained suspended, but fire extinguishing work was ongoing. Risk assessments were conducted and the mines were declared safe for essential services employees. All shafts had pump stations that had to be manned non-stop.

Multi-disciplinary task teams had been established to conduct audits in areas of reporting, ventilation, accumulation of gases, common central control room, re-entry conditions, personal protective equipment and emergency preparedness. Before any of the operations were started, a full risk assessment was conducted. The mine had intensified its safety control measures including a dedicated fire detection operator; control room register; additional training to technicians; review of fire alarm reporting structure; shaft entry authorisation; monthly systems compliance audit; and sufficient compressed air at refuge chambers at all times. There also would be an improvement on the self-rescuer kit.

Discussion
Mr C Huang (Cope) sought clarity on the main cause of death. He asked if the death was as a result of fire wounds or inhalation of noxious gases.

Mr M Sonto (ANC) was also interested in knowing the causes of death, as that would make the work of the Committee easier. He wanted to know the time it took for a noxious gas to incapacitate a human being. This was against the background of gas concentration levels that could sometimes be very high in mines. It was possible that the deceased could easily have been incapacitated before they could even reach for and put on their safety equipment.

Mr Robert replied that although the results of the final autopsy had not been released, there were no indication of burns on the bodies, and the deceased had been in an area with a high concentration of gases.

Mr Huang wanted to know if the high concentration of carbon dioxide was as a result of blasting that had happened earlier in the mine. He asked if there were any preventative measures to counter the uncontrolled spread of gases.

Mr Roebert replied that an investigation team had been assembled to look at the whole incident, and agreed to by all stakeholders and it would be led by DMR. The first task of the team would be to interview all of the witnesses - the 15 people who were rescued on that shift. These employees had been given time off from work, and interviews were scheduled to start on Monday 30 July. A lot of documentation had already been collected. Based on the concentration level of CO2 readings that were taken at the scene, it was unlikely that the concentration could have been as a result of blasting. It was definitely due to fire. The PPM system was the preventative measure used to pick up and sense the levels of concentration. The PPM picks up the concentration levels and individuals react accordingly.

The Chairperson interjected and sought clarity on the statement that the high concentration of gases and the fire could not have resulted from blasting. When one blasted, a measurable amount of gases were emitted; why would blasting be discounted as the cause for the high concentration levels of gases and the fire.

Mr Roebert replied that he could not answer that. The fire was still raging, and only once it had been extinguished could it be said with certainty as to what caused it. The fire was situated in an old working area that was last mined in 1999. Access to the area was a challenge and limited.

The Chairperson queried the accessibility to the area. That Goldfields was able to make gas readings at the mine indicated that the mine was accessible; therefore the cause of fire could have easily been ascertained.

Mr Roebert replied that the readings were taken by the rescue teams before the area was sealed off.

Mr Huang asked if communication at the mine was functioning and if it would have been easy for the deceased to communicate with managers about the challenges they encountered underground.

Mr Roebert replied that the bodies were found in the proximity of the refuge bay. The exact positions had not been marked off. It was unclear if the deceased had made it into the refuge bay or were still on their way. The investigation would determine the exact locations, and if they could have made it to the refuge bay.

Mr Huang interjected and wanted to know if the whereabouts of the refuge bay should not have been the first thing ascertained, instead of waiting for an investigation that could drag on for a long time.

Mr Roebert replied the rescue team had issued a plan that gave an indication of how far the deceased had been from the refuge bay - most of whom were close enough but only just.

Mr Huang sought clarity on the statement that all the deceased had their rescue kits on.

Mr Roebert replied that the deceased had the rescue kits, and three of them were supervisors who also had the detector kit.

The Chairperson interjected and said it was normal in these circumstances that the blame would be shifted to the deceased. The delegation was raising an issue of the deceased not donning their masks. He hoped the issue of not donning their mask was not intended to shift the blame. Could the presenter first answer an earlier question about the time it took for a person to be incapacitated after inhaling noxious gases.

Mr Roebert replied he was not a medical doctor and was thus not qualified to comment on the period of time it took a person to collapse after inhaling noxious gas. He knew a bit about the use of CO - mainly generated by cars - to commit suicide. The limit for CO2 at the mine was 30 ppm. This was a medical question.

Ms F Bikani (ANC) interjected and asked if high gas concentrations were not an expected occurrence when blasting took place, thereby warranting rescue teams to be on standby. She asked if managers were not alert to the eventuality of high gas concentrations, especially if 30ppm was the limit. Was there no research that could be done prior to blasting that could provide an estimation of the emissions?

Mr Roebert replied that there was a centralised blasting system for the mine. Normally the heads were all clear before the blast and no one was allowed underground except for essential services. There was a re-entry period only after three hours when people could enter the working place. As the gases left the mine, there would be clearer reading of the head and once it was clear, then the work started.

Mr Huang wanted to know the actions taken after the fire detection heads had given off an alarm soon after blasting. Were these alarms monitored, and what was the procedure that was normally followed in such situations. Why was so much time allowed to elapse after the first alarm had rang before rescue teams were called in?

The Chairperson asked how much time was allowed before the employees had been allowed into the mine.

Mr Roebert replied this would have taken place at 17h00, just over an hour before the first detection of fire.

Mr Sonto interjected and sought clarity on why the tramming crew - and not the rescue team - was allowed into the mine at about 20h00, if the fire was detected at about 18h45. Why would the normal workers be sent down? This piece of information was crucial and the Committee wanted clarity.

Mr Roebert replied that the heads gave off an alarm shortly after blasting, and cleared; but at 18h45 the heads gave off an alarm again, and they cleared again.

Mr Sonto interjected and wanted to know if, in the view of the mine, confusing signals given by the system meant it was safe to go underground and start operations.

Mr Roebert replied this was not the case. There was a procedure that needed to be followed by the operator, to call out the personnel, and determine if the head was faulty or the alarm had rung properly. This was not done in the day in question, but this formed part of the investigation.

Mr Huang voiced displeasure at the time that was allowed to lapse before the investigation could be undertaken. He accepted that it was humane on the part of the mine to allow workers time off. However, when those employees returned, much of the detail would be forgotten. This was a serious case and needed to be investigated immediately.

Mr Roebert replied that this was a dilemma, and the mine would have preferred it if the investigation was done immediately. The agreement with DMR and the unions was that the investigation would be done and led by DMR. Evidence was available.

Ms Bikani wanted to know the kind of risk assessment process that was in place, and what normally happened, before blasting.

Mr Roebert replied that the entire mine was cleared before blasting. The process would be conducted from the central blasting system. There would also be the fire controls systems that would check if the fire heads were clear for the three-hour re-entry period. The three-hour period also catered for the seismic window.

Mr Huang complained that the Committee was not getting a clear picture of what had actually happened. In these situations, the employer would usually go in immediately and investigate to mitigate the risk and protect other workers. What had happened?

Mr Paul Pretorius, Vice President Legal Services, Goldfields, said the company was required by law to go in immediately and investigate. And a report of such an investigation was then forwarded to DMR. The requirement was that the cause of the accident had to be established and strategies to mitigate future accidents would have to be proposed. Due to the complex nature of this accident, there was an agreement with DMR that DMR conduct this investigation. The accident area was still cordoned off, and the mine had put in place a set of remedial actions that had been discussed with DMR. These would ensure that such occurrences were countered in future.

Mr Lorimer wanted to know if it was appropriate to go on with the briefing if there was no official report at hand, and whilst DMR was still busy with the investigation.

The Chairperson replied that it was appropriate. The Committee had asked Goldfields to say what had happened so that it could determine the way forward. The Committee might decide to visit the mine or even bring DMR and the unions to Parliament. For now Members had to familiarise themselves with what had happened.

Mr Lorimer complained that the meeting was not achieving anything in ascertaining the facts of what had happened. He would have preferred to hear what DMR and the unions had to say.

The Chairperson replied Members were getting the information they required through asking detailed questions and that would give a sense of what should happen next in relation to this matter. The questions asked by Members were leading the Committee in the right direction. He allowed the delegation to continue answering questions.

Mr Huang requested that Members be taken step by step through what had happened.

Mr Roebert replied that on the day in question there was a normal shift underground. Everyone was taken out of the mine in preparation for blasting, and it took place at 17h00. On Saturdays there were no normal nightshifts. There were certain exemptions with the Sunday labour permission that on an overtime-voluntary shift there would be people going down, to perform certain work. This was an approved volunteer shift and supervision was in place. The crew was going to go down to the Level 39 working place to load the mud, inspect and fix the rails, and clean the tunnel. These were the tasks given to the crew, and they were allowed to proceed at about 20h00. There was an alarm in the control room. That alert was not communicated to senior managers by the person responsible for the control room on that day.

Failure to report the alarm would be the subject of the investigation as well. The team was allowed into Level 39 where they were expected to perform the work. Much later there were reports of people collapsing and it would not have been that easy to see them. People fell at various positions, and it was possible that other crew members were far behind the deceased; or it could be that they tried to help, but could not cope with gas, and eventually retreated to places of safety. The survivors were all taken to hospital and were placed under observation but later discharged.

Mr Huang asked if employees were adequately trained in using and reading the self-rescue equipment, and also if they knew how to access the refuge bay. What might have occurred in this situation and how could it be prevented in future?

Mr E Mtshali asked if the mine conducted mock rescue operations to help employees better understand the equipment.

Mr Roebert replied in the affirmative, especially from the perspective of injuries. Every month a rescue drill was performed, where supervisors showed the teams how to access the refuge bay and how it functioned.

He did not know what might have happened in this situation, but this was the kind of information that the investigation team would have to provide. Four of the deceased had GDIs with them. The investigation would also determine if the GDI alarms had worked or if they were faulty. That would be subject to the evidence provided by the witnesses. Initial investigations showed that three of the GDI alarms had, indeed, rung at about the same time. This information was downloadable on the computer because this piece of equipment operated more like a “black box”.

The Chairperson asked if the mine thought it was erroneous to send the crew down despite the alarm having gone off. He asked if the blame was with the official controlling the system.

Mr Roebert agreed that it was wrong to send the crew down when the alarm had been set off.

Mr Sonto wanted an update on the fire at the mine and why it took so long to extinguish.

Mr Roebert replied the fire was still going on.

KDC fire strategy
Mr Roebert said the mine had not had a fire for the past four years except for smaller outbursts here and there. When the mine fought fire, it first removed all persons from any area that could be contaminated. But also closed off and sealed any workings on adjacent shafts that could be affected and in this instance, these were Hlanganani and Masakhane shafts. These shafts could be affected but had been closed off; no work was happening there.

The next step was to establish a chimney to channel smoke. If reconnoitering the area was not possible, there were installations of water spray pipes on the walls. Sealing off appropriate areas to limit the feed of oxygen (O2) was another contemplated method.

A seismic ring of 1km radius had been established around the fire area, and there was no mining happening even in nearby mines. The tendency had been that when the fire was burning it destroyed the supports in the mine; and when it was extinguished, one was left with no supports. This condition could lead to rock falls and further accidents. Emergency and fire teams could not access these areas whilst the fire was burning because there would be no supports.

At Ya Rona (4) Shaft a fire control room had been established on surface and was being manned day and night by senior managers, engineers and environmental and safety officers, and also mine rescue services. A scribe was appointed to record the sequence of all events.

As pertaining to the Ya Rona fire, a reverse chimney was established to channel down the air to Level 39. While the chimney was being established, access could not be gained due to high temperature. Another challenge in the area was visibility. The chimney was being monitored day and night.

Mr Roebert said it was hoped the fire would be extinguished in two weeks, as gas levels were declining and the fire was contained. All the Medical Rescue Services (MRS) teams were fully trained and equipped to deal with the conditions.

Discussion
Ms Bikani asked for a statistical breakdown of fatalities at Goldfields mines.

Mr Roebert replied in 2007 there were 34; 2008 (20); 2009 (25); 20010 (17); 2011 (18); and 2012 (12 so far).

Mr Sonto said the statistics were a cause for concern. He asked what were the contributing factors to the high rate of fatalities at Goldfields mines. Every year there had to be families crying.

Mr Roebert replied that fatalities used to be due to fall of ground and seismic events. Deaths as a result of gas had last occurred four years ago. But there was a safety strategy that had been put in place, and it involved overhauling and rolling out safety systems around the transport system and occupational health. The company had also done away with nightshift on many of Goldfields shafts. There also would be leadership and behavourial change programmes. Intense workplace audits would also be conducted. Every working place in the mine would be visited every 21 days. There also would be a review of the bonus system, and it would be awarded on the basis of sound and compliant safety measures.

Mr Sonto said he understood all the precautionary measures but he still believed that the deceased were overcome and incapacitated by gas.

Mr L Gololo (ANC) wanted to know if illegal mining was not a challenge for Goldfields, and if there was a possibility that illegal miners might have caused the fire, especially as they take their cooking equipment underground.

Mr Roebert replied illegal mining was a challenge but there were dedicated teams that chased and arrested illegal miners. It could not be ruled out that illegal miners might have caused the fire but such a determination would be made by the investigation. But, yes, there was the challenge of illegal mining as experienced by any gold mining company, and these were reported to DMR. Goldfields protection services worked hand in hand with the police on this phenomenon.

Mr Gololo sought clarity on the compensation for families.

Mr Roebert replied that a death benefit was paid out to all of the families. The Chief Executive Officer had also initiated a plan to take care of the educational needs of all the children of the deceased. On top of all this, Goldfields allowed a family member to come and work in the mine to replace the one who had passed away.

Ms Bikani commented that not enough was done to minimise fatalities and a strategy was needed to curb the rate at which fatalities were occurring. A detailed strategy would be appreciated by the Committee to really convince Members that something tangible was being done.

Mr Roebert welcomed the comment and said the message would be conveyed to stakeholders in the August tripartite session (with DMR and the unions).

Closing remarks
The Chairperson thanked Goldfields and said there were areas that needed further clarification. The company might be invited back to Parliament, but the Committee would allow for the completion of the investigation. It was a very serious concern when dealing with loss of life. It did not matter if it was caused by fall of ground, moving machinery or even occupational diseases. The level of fatalities in the mining sector was a cause for concern. The Committee was happy to note that there had been a decrease when compared to days gone by. Depending on what the investigation yielded, there might be a need for senior managers to take full responsibility for what had happened. The Committee might want to visit the mine when the fire had died down.

The meeting was adjourned.