As open pit and underground mines become deeper, and development and operations of mines becomes more hazardous, the entry of automation introduces unexplored territory, writes Jared Haube.
Have we actually given enough thought as to what risks have emerged as a result of rapid advances in automation technology?
I spoke with Raj Sreenevasan, Principal Instrument Engineer and automation expert about a unique field of study in mining that has the potential to significantly influence the impact of automation.
According to Raj, the mining industry needs to learn from historical near-misses and incidents prior to deployment of automated operations; in context, reflecting on black swan and white swan events. Raj referred to a book called “The Black Swan – The pact of the Highly Improbable” by Nassim Nicholas Taleb, which explains that a black swan is an outlier event; in that it remains external to the domain of regular expectations, because there aren’t any indicators in the past that could identify the event’s possibility of occurring. It also brings an extreme impact or consequence.
“People who attempt to evaluate risk must understand that black swan events cannot be predicted reliably using current risk analysis tools,” Raj said.
One black swan event in mining to which Raj referred, included the Aurul Gold Plant in 2000, in which a dam had burst near Baia Mare, Hungary, releasing 100,000 cubic metres of cyanide-contaminated water into farms and the Somes River. The Hungarian Government expressed multiple condemnations, citing that it was ludicrous to store cyanide next to a river in the first place. Different variables for the cause were discussed; from extreme snowfalls to anomalies in the dam design. The unpredicted occurrence left a disastrous impression on the environment, and these two factors define the essence of a black swan event.
Introducing large scale automation to mine sites will present its own portfolio of unpredictable possibilities, because automation is still in its early stages as an innovative industry application; this also limits the opportunity for white swans to prosper.
Raj went on to explain that a white swan event “exists through the consistent awareness of people; because as long as its causes are discussed and new methods of prevention are developed, it is a white swan event. It lives in memory.”
Intriguingly, Raj indicated that these white swan events have the potential to become grey over time.
“When people forget the lessons learned, and the consistency is broken, white swans can go grey over time. Take the three coal mine disasters in Moura, Queensland, after the 1994 disaster, the Queensland government had had enough and amended the Mines Safety Act, when the Trigger Action Response Plans – TARPs – became mandatory for coal mines in Australia. However, the Pike River Coal Mine disaster in New Zealand showed that the white swan had indeed turned grey.” The 1994 disaster near Moura, which Raj mentions, was indeed disconcertingly similar to the Pike River Coal Mine incident.
In the Moura disaster, an explosion occurred in the No. 2 coal mine on 7 August 1994, after which rescue attempts had been organised; but not before a second explosion 18 hours later rendered search efforts obsolete. Eleven miners died.
The Pike River Mine incident involved an initial explosion which killed 29 people on 19 November 2010, and three subsequent explosions prevented any rescue attempts, eventually resulting in the mine being sealed.
Raj’s example points to the TARPs in Australia as the white swan application to the event, and the transition to grey is presented through the occurrence of the Pike River Mine explosions, fourteen years after the last disaster in Moura.
Although it is, as aforementioned, too early to analyse white swan events in automation, Raj’s observation paints a clear picture of the importance of consistent awareness.
“Introducing large scale automation to mine sites will present its own portfolio of unpredictable possibilities, because automation is still in its early stages as an innovative industry application…”
Some successful lessons learnt in the mining industry, which influenced legislation, include:
- The Gretley Colliery disaster on 14 November, 1996, in NSW, where four men drowned from a sudden inrush of water originating in the abandoned workings of the Young Wallsend Colliery. This event led to an update of the NSW Mine Safety Act (targeting an introduction of industrial manslaughter legislation in State Parliament).
- The Esso Longford gas explosion on 25 September, 1998, in Victoria, where two men died due to a rupture in a heat exchanger. The event prompted the approval of the Victorian Major Hazard Facility legislation (the national MHF legislation follows the Victorian model).
As the Australian mining sector readies itself to cross the threshold to mine site automation, Raj believes that risk probability will need to be explored to maximise preparations for the unexpected.
“Although it’s quite early to apply the theories behind black swan and white swan events for automation, it’s important that there’s strong investment into mitigating risk potentials,” he said.
Dependency on technology is not a new phenomenon, and the rapid advances are transforming not only the way in which companies conduct operations; but also how industry professionals think, and that’s why investing into the awareness of swans could have significant influence in the long term success of automated mine sites.
WHY INTEGRATING SIMULATION SYSTEMS IS CRUCIAL TO AUTOMATED MINE SITES
Mining and resources companies are investing in automation worldwide, in pursuit of optimising operational efficiency, reducing expenditure and increasing productivity gains.
The transition from manual to automation exposes a “twilight” period; in effect, the stage where staff require new skills and professional development to meet the technical requirements of automation management, and new safety procedures.
I spoke with Julian Jones, another automation specialist about how simulation in automation systems could be the key to innovation for facilitating staff skill development in the age of automation.
“Another advantage lies in the fact that you have the ability to train for rare yet catastrophic incidents; a bit like a plane landing without its wheels down.”
Can you talk about some of the inherent advantages of integrating a training simulation system?
There are many key advantages which reflect the flexibility of an integrated simulation system.
You’re not training in an offline environment, so production and equipment risks are effectively eliminated.
Another advantage lies in the fact that you have the ability to train for rare yet catastrophic incidents; a bit like a plane landing without its wheels down. We can set the scenarios up to mitigate the risks in case they ever do occur, training the staff to prepare them for those rare events. It’s almost impossible to train them in real plants; you wouldn’t set it up to be an unsafe environment!
We can also get definitive competencies for a range of scenarios which might take years to occur in the real world. Lots of things happen in plants and mine sites, so you want to make sure your operators know how to react to those scenarios; to ensure that they are appropriately familiarised with the reactionary procedures.
Speed of uptake is another advantage; engineers can use simulation for optimisation. Put into context, think of a pilot undergoing training in a simulator, because it’s exactly the same. All the benefits of him or her doing that in a simulator, instead of a planeload of passengers; all the same benefits apply.
What are some of the critical success factors to integrating simulations with automated systems?
So, like most things, the prime critical success factor is understanding what the client wants to get out of the system. If the client wants his or her operators to understand the basic navigation and user interface, for example, then you only need a very basic level of simulation, and you spend that money accordingly.
If the client wants to be able to go right up to the engineers to use it for optimisation, then the fidelity and level of simulation needs to be a lot higher.
Other success factors include ensuring you have a good understanding of how the plant works. Spend a little money, you get a little bit of simulation fidelity; spend a lot of money and you get the Matrix essentially! But ultimately, understanding how the plant works in real life.
Can you talk about integrating immersive technology into a simulation environment?
There is a range of technologies which are readily available for deployment into a simulation system, and for a range of benefits.
One benefit might be to show the operator what the actual plant looks like in the configuration the simulation is currently at; such as equipment positions, how much dust is in the plant, what the steam levels are, and even how noisy the plant is.
One thing we’ve done recently is integrate virtual CCTV into the experience so that virtual cameras show the plant in the correct state, and replicate the view of the actual cameras that the operators have. There’s a range of immersive technologies for integration; but essentially, it’s around giving more context as to what the physical asset would look like in that particular scenario or state.
What kind of roles can benefit from integrated simulation systems?
Operators, engineers, maintainers, planners – it all depends on what you simulate and to what level of fidelity. Our focus is on plant control room operators – allowing them to get the most production from an asset safely.
Any role that needs to look at a scenario that would be difficult to set up in real life can benefit.
Are there inherent challenges facing staff for simulation training?
A challenge we often see is from older plants where the information is not readily accessible; simple things like how long a conveyor is, or how long a valve takes to open or close. It also becomes a long and laborious process.
With newer plants it’s not a problem; you go to a drawing or a data sheet, and it’s a seamless process. However, with older plants, some of them have been there since the ‘70s, and that makes it difficult to access important physical information.
Another challenge relates to variability of the complexity of the plant being simulated. An iron ore plant would be simple, for example, because there aren’t many complex reactions or processes occurring. However, when you start going into heavy process industries, that’s where it gets complicated.
The level of fidelity required is another influential factor; looking at the degree to which a simulation requires for recreating the actual plant or mine site. Some simulations require a much higher level of fidelity.
Is there a high market demand for these simulation systems in Australia?
Traditionally the power station industry was the mainstay for this sort of technology; however, now mining is demanding this capability. The operators at the end of the day are running the asset – sure, the control system is making the second by second decisions – but the operator needs to be able to get the most out of those systems. If the assets are worth $1Billion+ each, the investment in the person running that asset day to day is seen as having a great business case. So the answer is yes, and it’s growing.
SOURCE
This article has been republished with kind permission of Mining IQ. Visit their website, www.miningiq.com, for information on a range conventions and seminars on automation as well as many other topics important to the mining industry.
“Not everyone was aware that CO make was the new standard, or knew how to calculate it.”
Add Comment