Hide this alert
InfoMine Home

Risk Assessment in Mining 

 
Author:Jack Caldwell (Robertson GeoConsultants)

Revised: June 2013

Summary

This review gives sources of information to help you undertake risk assessment for a project. Risk assessment is defined and its applications are outlined. Links to online courses, software, consultants and publications related to risk assessment are also given.

INTRODUCTION

Georgius Agricola (24 March 1494 - 21 November 1555, was a German scholar and scientist - known as "the father of mineralogy") says this about the risk of mining:

The critics say further that mining is a perilous occupation to pursue because the miners are sometimes killed by the pestilential air which they breathe; sometimes their lungs rot away; sometimes the men perish by being crushed in masses of rock; sometimes falling from ladders into the shafts, they break their arms, legs, or necks; and it is added, there is no compensation which should be thought great enough to equalize the extreme dangers to safety and life.

From the decision to invest in a new mine to the decision to go underground to work in an operating mine, hazards and resulting risks have to be evaluated. I know of no work (book, paper, conference proceeding, thesis) that adequately covers all aspects of mine risk in a logical and consistent way. And I do not try to rectify that gap here.

In this review, I point you to sources of information that may help you undertake your own risk assessment of your own sphere of operations at your mine. This scope recognizes that each risk assessment is unique; no two are the same; and a successful risk assessment leading to a sensible and sustainable mitigation/reduction of risk is only as good as the knowledge, skill, and effort of those undertaking the assessment.

DEFINITION

The term risk means different things to different people. The need for standards definitions became clear two decades ago when IUGS wrote the first risk glossary applicable to geotechnical and geological risks. General glossaries exist, like, for example the ISO 31000 one. The most commonly accepted technical risk definition is R=p*C where p is the probability or likelihood of an event and C is its consequence. If C is positive (a gain, profit), then the risk is an upside one; if C is negative (loss), then the risk is a downside one. Generally speaking Risk Assessments look at downside risks.

To take Agricola's example of falling off ladders: if ten people a day climb a ladder in a mine, for say 300 days a year, and on average one person per year falls off a ladder and breaks his leg, the empirical frequency is easily calculated to be one in about three-thousand per annum that climbing that mine's ladder will have the consequence of break your leg. As 1/3,000 is a small value, then we can consider that the empirical frequency and the probability p are equal.This is an example of objective risk quantification: numbers are used after the fact to calculate the "historic" risk.

Conversely, as an example of subjective risk estimation, assume you are trying to set your salary so that you are reasonably compensated for the risk of being crushed in masses of rock. There are many equations you could use, but none mean anything unless you and your employer can agree on the future probability of bone-crushing events. You will, no doubt, propose a higher number than the employer who will proclaim, that it has not happened before at his mine and therefore the risk is miniscule. You, on the other hand, may point to industry statistics and your subjective fears and judgment, using the argument that the future does not, generally, equate the past. It could get complicated.

But at some point you will have to agree that the risk boils down to the probability of a rock fall at a mine in general, the probability of the fall being at the mine where you work, the probability that you will be there when it happens, and the probability of you actually being crushed as compared to merely scratched.

THE BASICS

The basic steps in a risk-based decision making undertaking include:
  • Define Context: the social, organizations, political, and technical issues and environment
  • Define System: the components and their interactions
  • Define Risk Scenarios: what can go wrong; how can it happen; what controls are in place?
    • Estimate Consequences: who could get hurt; what facilities could get damaged; what are the costs to clean-up?

    • NB: consequences, like probabilities should never be taken as "one number"...there are too many uncertainties to allow that.
    • Estimate Probability: objectively from data or subjectively from deliberations,
  • Undertake Risk Assessment: "Pair" probabilities and consequences, compare them to risk tolerability/acceptability criteria in order to enable rational decision making on risk mitigation
  • Implement Risk Mitigation Measures: fix things; change systems; be prepared.

EDUMINE

I have a small space to cover this vast topic. Much more is available at EduMine on the topic. Their courses include:

SOFTWARE

Before we get into a subjective overview of different risk assessment packages, this wikipedia link provides a cold hearted run-down of the features, technical specifications etc. provided by each of the different software packages.

The best software I know of for supporting the quantification of the engineering and technical risk is put out by Palisade. I particularly like their code @RISK. I recommend you get a copy, read the accompanying book and use the software. I go so far as to opine that if you do anything less you run the risk (unquantifiable) of being judge unreasonable and unprofessional. (By way of disclosure, Palisade did not pay me to say this; they did send me a free copy of the code and I have enjoyed playing with it. If you send me a copy of your analogous code, and I like it, I may write similar.)

My other favorite is GoldSim Technology Group. This is how they describe what you can do with their software. "GoldSimsupports decision and risk analysis bysimulating future performance while quantitatively representingthe uncertainty and risks inherent in all complex systems." Go to their site to read more about the application of their codes in these aspects of mining risk evaluation:

NOWECO sells RiskDecision which is billed as Risk Management Software. At this link is their Overview. It is worth reading if only as a primer on the basics of risk assessment, and potentially as an inducement to purchase the software.

Vose Software provides three excel based products with varying degrees of expense and complexity. These include:

  • ModelRisk Standard (STD): Which provides "provides advanced Monte Carlo simulation in Excel", is free to download from their website.
  • ModelRisk Professional (PRO): This is the next level up and "adds 'objects' to ModelRisk Standard that greatly extend and simplify what you can model" and also includes tools for optimization.
  • ModelRisk Industrial (IND): Is the high end risk analysis product, providing advanced features not included in the PRO version including the use of Oridary Differential Equations and Database Connectivity to help solve industry specific problems.

FTA-Pro produces a computer code to help you compile fault trees, that old staple of a comprehensive of risk assessment. Their description of fault tree analysis is instructive:

Fault Tree Analysis (FTA) is a top-down approach to failure analysis, starting with a potential undesirable event called a top event, and then determining all the ways in which it can occur. In other words, Fault Tree Analysis is a graphical representation of the logical structure displaying the relationship between an undesired potential event (top event) and all its probable causes. Based on the FTA, mitigation measures can be developed to minimize the probability of the undesired event.

Fault tree software is also available from Relex Software Corporation.

The Intaver Institute produces project management risk software. This enables you to incorporate risk (random events & the chain of consequences) into your project scheduling. They use events chain methodology to do it; this is described thus:

Event chain methodology is the next advance beyond Critical Path Method and Critical Chain Project Management. Event chain methodology allows you to:

  • Define your project schedules with risks events and uncertainties while taking into account the moment of risk, relationship between risks, and other factors.
  • Visualize risks events on your project schedules.
  • Analyze your project schedule: compare main project parameters with and without risks, identify crucial tasks, critical risks, project success rate, and calculate chance of project completion.
  • Perform adaptive project management: track your project's performance with risks and uncertainties.
  • Calculate realistic buffers for critical chain project management.

Syntex Management Systems call themselves "global leaders in enterprise risk management software." I find it nearly impossible to work out from their copious words what their software actually does. It seems to me it enables you to collect data, sort and collate data, present events and incidents for consideration, and hence manage things that might go wrong, i.e., risks. Here is how they describe it.

Impact Enterprise: Facilitates the discovery and removal of exposures to risk that result in organizational loss; Tracks incidents, investigations and responsibilities; Encourages the creation of remedies, for pro-active and reactive situations; Facilitates assessment of corrective actions; Provides site-level and enterprise-level views of performance.
A word to conclude: make sure you read an understand the words of caution at the beginning of this section and the read the CAVEAT below.

CAVEAT

The following was probably written by the corporate lawyer for Palisades. But it acts as a sober reminder to those too infused with enthusiasm for the numbers from their computer printout on risk assessments.

Quantitative analysis techniques have gained a great deal of popularity with decision-makers and analysts in recent years. Unfortunately, many people have mistakenly assumed that these techniques are magic "black boxes" that unequivocally arrive at the correct answer or decision. No technique, including those used by @RISK, can make that claim. These techniques are tool that can be used to help make decisions and arrive at solutions. Like any tools, they can be used to good advantage by skilled practitioners, or they can be used to create havoc in the hands of the unskilled. In the context of Risk Analysis, quantitative tools should never be used as a replacement for personal judgment.

CONSULTANTS

You know that the principles and application of risk in mining are widespread when you get nearly five hundred consultants listed in ConsultantMine for the keyword risk. Personally I would first contact Oboni Riskope Associates and then do what they advised. (Again by way of disclosure, I have heard of their principal but never met him. I judge on the basis of the website information.) Other specialists listing risk as a resume keyword may be accessed via this link.

TECHNICAL PUBLICATIONS

Available from the SME bookstore is a great little book entitled Sustainable Mining Practices. Its authors are Vasudevan Rajaram, Subijoy Dutta, and Krishna Parameswaran. In a chapter called Tailings Management by Robert E. Melchers are nine pages of text on the application of risk assessment to tailings impoundments. This is the most succinct and informative treatment of the topic I have come across in a while. Go take a look before you undertake a risk assessment at your mine, at least at your tailings impoundment.

I cannot access the AusIMM publications from 2003: Mining Risk Management. But you can get it for $90 at this link. Individual papers are available for purchase at $15 each. This is crazy-they should put them in with OneMine and be done with it. OneMine has over 900 technical papers related to the mining industry with the word risk in the title or contents. The papers cover everything at a mine from conveyor belts to environmental sampling. Well worth the cost of a membership of SME.

In the InfoMine library are over 100 publications listed for keyword risk. The latest from 2007 is entitled Hedging, Whose Risk? To the earliest from 1987 A Probabilistic Approach to the Long-term Stability of Uranium Mill Tailings Impoundments. Which gives me the opening I need to emphasize that there are certain things you should not do: one of them is apply probability to the failure of closed uranium mill tailings impoundment. No matter, this particular paper was written before my colleagues and I fixed the error of their ways on the UMTRA Project.

CAVEAT

The following was probably written by the corporate lawyer for Palisades. But it acts as a sober reminder to those too infused with enthusiasm for the numbers from their computer printout on risk assessments.
Quantitative analysis techniques have gained a great deal of popularity with decision-makers and analysts in recent years. Unfortunately, many people have mistakenly assumed that these techniques are magic "black boxes" that unequivocally arrive at the correct answer or decision. No technique, including those used by @RISK, can make that claim. These techniques are tool that can be used to help make decisions and arrive at solutions. Like any tools, they can be used to good advantage by skilled practitioners, or they can be used to create havoc in the hands of the unskilled. In the context of Risk Analysis, quantitative tools should never be used as a replacement for personal judgment.

APPLICATIONS

GENERAL
In this section let us discuss application of the ideas and principles of risk assessment as described above and in the links provided. We do this to set out ideas not readily found. Hopefully these ideas will help you if your boss comes in late on a Friday afternoon with the good news that you have been selected to lead a detailed risk assessment of a critical facility on your mine.

MINE WATER BALANCE
We start with the mine water and chemical mass balance. If you have too much water or too many contaminants in the water, you may not be able to comply with your mine's discharge criteria. If you have too little water, you may not be able to run your mine, keep the process plant operating, transport the tailings to the impoundment, or irrigate newly planted reclamation vegetation. All good reasons for quantifying the risk of too much or too little and go into a Risk Based Decision Making process.

First step, as in all risk assessments, is to understand what you have got. Quantify the mine site and surroundings surface water conditions. Quantify the mine site and surrounding groundwater conditions. Then compile a mine water and chemical mass balance. This may be as simple as a sketch, more detailed as an Excel spread sheet, or as complex as a GoldSim computer simulation.

The second step is to establish the probability distribution functions of important parameters like precipitation, evaporation, seepage from the tailings impoundment, seepage into the open pit or shaft, and the need for more or less process water. In many cases you will not have enough data to support this, hence you will rely on point estimates methods and certainly not go through with a nuclear missile (Monte Carlo) to kill a fly (your poor data).

The same process should be implemented for production of acid rock drainage (ARD) from the pit walls, the waste rock dump, and/or the tailings impoundment. If there are other contaminants that may bedevil compliance with discharge criteria, establish the probability distribution functions for sources of these.

The third step is to analyze the situation. If you are using a GoldSim model this information may be directly input, the correct subroutine called up, Monte Carlo made do its stuff, and you will get a printout of the probability distribution curve of whatever parameter you selected: generally discharge quantities from the site of quantities of fresh water make-up required to keep the plant going.

Remember that Monte Carlo simulations results are influenced by the mistakes you made in selecting the distributions of your parameters. You may be better of working with point estimates methods, or consider both ways and use good sense before you decide which results to adopt.

If you are smart, you may be able to program your Excel spreadsheet to do this.

Things may get more complicated when your boss asks questions like:

  • Did you take into account the demand on our offsite supply by the local village?
  • Did you take into account the integrity of the supply pipeline from the well field-by considering things like lightening, sabotage, and rust-induced failure?
  • Did you consider the possibility of local groundwater table recovery during very wet years leading to increases in acidic seepage to the pit?
  • Did you consider the chance that winter freezing of the tailings will entrain lots more water than we estimate?

There are literally thousands of such questions that can and probably (a risk!) will be asked as the budget diminishes and time (schedule) drags on. Now is time to call in more folk-people with detailed knowledge and information you may not have at your finger tips. Invoke the many techniques of Decision Making. Set them to work in a Value Engineering session. Compile fishbone diagrams linking events, and causes & effects. Prepare fault trees: those diagrams that show one event and subsequent events; and ascribe a probability of occurrence at each node. Before you get in a "paralysis by analysis" though, remember to look at what risks are bearable/tolerable and concentrate on those that aren't.

Now you will need a code like @RISK to do the number crunching-that is unless you have a very simple situation and a fast finger on the calculator.

Once you have the numbers and your boss's understanding and concurrence, (you will not get that if you do not master the arto of presenting risks and packaging the right information)formulate a plan of action to reduce the occurrence of the undesirable outcomes. Remember, zero risk does not exist, so prepare for the worse, and never stop thinking and planning for possible crises. Communication plans and monitoring (observational method) will help.

TAILINGS IMPOUNDMENT FAILURE
This is a favorite topic of technical papers-you should be able to find a few in those places I list above. But whatever papers you find, I guarantee they will not fully reflect the situation at your impoundment. As always, you will have to fully define and characterize relevant conditions at your impoundment.

In Sustainable Mining Practices, these tailings impoundment failure causes are listed: weak foundation; liquefaction of tailings; overtopping; piping; slope instability. There are others including erosion, washout, excess pore pressure build-up, burrowing animals, and excess seepage. You will have to decide on a case-by-case basis what is significant at your site. Never rely exclusively on this type of list.

MINE CLOSURE PLANNING
You are told to establish a budget for mine closure ten years hence. You are told that it is OK to spend a little money each year until then to get ready, but preferably you should defer expenditure until closure-and then spend money only to "equalize" the risk of failure of all post-closure facilities.

Sort of like Henry Ford who sent inspectors to the junk yards to find which components of discarded Model Ts were still intact. Then he reduced the quality of those still-intact components, reasoning that there was no point of spending more to make something last longer than the basic car itself. This approach prevailed in the auto industry until the Japanese arrived and look where we are now: I am on my fourth Honda. I confess my kids love the Fords, owning between them a Focus, an F150, and an Explorer. My ex has a red Mustang, so maybe it's just me? And your boss is right to set this technical approach to closure.

You have to get the answer. To do so will probably involve compiling a fault tree for each post-closure facility, identifying potential failure modes, costing avoidance works, and adjusting designs and failure trees until you get that perfect solution: spend the least possible amount at closure to put each in a state where its probability of failure is more or less equal to that of the other closure facilities.

Things may get more complicated if you also have to consider the physical and cost consequences of failure. Good Luck.

| Back To Top |

News

Careers

Suppliers

Consultants


Publications

EduMine Courses

Events

Links

Publications Search