Why Gold Mines Fail

It has become fashionable to blame the flow-through tax rules for the excesses that led to the mine development failures of the 1980s. But this is not entirely fair. My colleagues and I were in the unique position of closely observing some 25 advanced gold projects in the late ’80s. Only three succeeded and lived up to expectations. The rest, despite some considerable capital outlays, can be considered, at best, ill-advised investments. Technical weaknesses within certain industry sectors were the prime cause, and artificial time constraints in the flow-through regulations merely exacerbated these problems.

All of these failures came about because the unit cost of production was too high: it cost more to produce an ounce of gold than that ounce could be sold for on the market. While this sounds simple enough, one has to ask what happened and why. Why were there so many problems? Why were so many people wrong so often?

The answer to the first question is relatively easy and can, in most cases, be reduced to two items:

* lower mill head grade than expected; and

* lower mine productivity than expected.

Rarely do metallurgical and mill design problems kill a mine. Usually, such problems are symptomatic of more serious deficiencies elsewhere in the project. Although many metallurgists would disagree, there is, in most mining situations, considerably less risk in designing and operating a plant than there is in designing and operating a mine. Even if there are early metallurgical problems, they can usually be solved if the natural resource is available in the mine.

In this paper, we speak of mill head grade as opposed to ore reserve grade, because the term “ore reserve” means many different things, all of which have been well documented — mineral inventory; mineral resource; resource inventory; geological reserves; drill-indicated reserves; minable reserves; proven, inferred, probable, possible, indicated, measured and (my own favorite) minable geological inventory. Each person who uses one of these terms, we hope, understands what they mean. But despite the efforts to improve the situation, confusion still reigns.

We don’t advocate abandoning this terminology, but refer directly to the mill head grade because it is a good catch-all to account for the two main bogeymen of gold ore reserve estimation: dilution and nugget effect. About these, more later.

Mine production capability is an often-overlooked or trivialized factor that can be as important as the head grade. While the head grade controls the revenue side of the equation, productivity is the major factor on the cost side. It is our observation that operating cost estimates for most projects are usually based on some logic, provided they have been developed from a zero-cost base rather than a unit-cost base. The process generally falls apart when determining what number to use as a divisor — how many tonnes per day the mine can produce. The important factors in this determination are vein width, vein dip, continuity, wall rock conditions, number of workplaces, mining method and infrastructure.

These factors may seem simplistic to experienced planners, but some people have forgotten, ignored or never knew about them. (How else can we explain a 1,000-tonne-per-day mill sitting on top of a 500-tonne-per-day mine, as we saw at one project?)

* Vein Width — Perhaps the most obvious factor and missed by very few. Obviously in cyclic operation, the more volume extracted per cycle makes the operation more productive.

* Vein Dip — Who has not worked on a project where the vein dip was 30deg to 40deg (a situation where gravity is more a hindrance than a help)? If you are going to have to carry every tonne of ore out of the mine, make sure you allow for it.

* Continuity — Structural disruption within a vein must first be understood, which means adequate drilling and development; 25-ft centres used to be the standard, but how often is that the case now?

* Wall Rock Conditions — How much unproductive time needs to be spent on ground control?

* Number of Workplaces — This is a key factor in all planning exercises. Determining the number of workplaces a mine can support is controlled by the service infrastructure and the production cycle. Can we physically move ore and waste out of — and materials, fill and men into — the mine?

Because this is something of an art (i.e., usually a result of calculations tempered with a “gut feeling” for the situation), it can only come from experience. Too often, this part of the planning is left to the most junior members of the technical staff.

* Mining Method — We keep reiterating what may seem very obvious to a lot of people — in this instance, the importance of choosing the proper mining method. What could be simpler, especially in vein gold mining where, with few exceptions, it is either cut-and-fill or shrinkage? How, then, can we explain people trying to mine a 75-cm vein with 2-boom jumbos and 5-cubic-yard load-haul-dump machines? The operation ends up with a very respectable operating cost per tonne, but what about cost per ounce?

* Production Infrastructure — How easily can we move ore and waste through the system to the mill? Are there bottlenecks? How much rehandling is there? How much storage is available before and after crushing?

Mill head grade and mine productivity are not independent of each other. The following sequence has happened time and again and is usually the first step down the road to failure: the grade drops, the mine manager screams for more ore, pressure is on the front-line supervisor, everything goes in the ore pass, tonnes go up, grade goes down, cost per tonne goes down, cost per ounce goes up. In the late 1970s, at one of Canada’s best known gold mines, a plan to lower the unit cost by longholing backfired when the grade dropped too low. Management wanted more tonnes, the grade dropped more and the tail-chasing began. The operation was fortunate because a minable resource still existed and management corrected the problem by returning to the proper mining method.

Having shed some light on the first question of what happened, we now turn to the second question — why? From our observations, a number of common factors have emerged and they are all traceable to inexperienced mining professionals. Most severe perhaps is the lack of true mine geologists in many of these projects and the related lack of understanding that most exploration geologists cannot suddenly switch hats and become mine geologists. This problem has manifested itself in inaccurate ore reserve studies and production decisions being made on the basis of what are really only geological reserves.

A proper ore reserve is impossible without truly understanding the way a stope operates. Most exploration geologists think on a broad scale and in general terms. A mine geologist thinks on a much smaller and more detailed scale. He pays attention to small geological details. For example, 20% dilution sounds like a lot in a 1.5-metre-wide stope, but in reality it is only 1 cm (or six inches) off each wall. The ore reserve estimator must understand this.

Similarly, mining engineers at the junior and middle management levels these days are products of the late 1960s and ’70s. In earlier days, all engineers at some time worked in gold mines, large and small. With the closure of numerous large old gold mines in the 1950s, ’60s and ’70s, many of today’s engineers (and geologists) did not have the opportunity to spend time in these valuable training grounds. And with the 1960s came the transition to mechanization — headings were 4×5 metres and the future was ST-8 load-haul-dump machines. Everyone was working in the large mines where 2,000 tonnes per day was nothing.

Suddenly, along came the 1980s and every new project was a 50-cm vein, dipping at 35deg with 200,000 tonnes of reserves at six grams per tonne, and access through a 2-compartment shaft dating from the 1940s. To say that technical people were ill-prepared is an understatement. Except for the part about the care and
feeding of horses underground, the Peele handbook suddenly became the modern reference again. Compounding the inexperience, too many consultants were too eager to please their clients and too few operations people were involved in the go-ahead decisions on mining projects.

Unfortunately, there will always be independent consultants who do not have the internal fortitude or discipline to tell a client what he should be told, rather than what the client wants to hear. Similarly, accountants and lawyers seem more and more to be in decision-making seats these days. Many prospectors and promoters, although absolutely vital to the growth and success of our industry, are not technically equipped to prepare feasibility studies and properly judge their conclusions. There will always be mine failures, but the chance of failures can be reduced by following a systematic, logical approach to analyzing each mineral occurrence. Some may find the following list of dos and don’ts helpful. It is neither definitive nor comprehensive; nor is it meant to point any fingers at anyone. The list is applicable to vein gold deposits and any other mineral resource.

Dos

Do proper data collection and analysis, including comprehensive sampling, mapping and assaying. Sample in a consistent manner, suitable to the ore in question. (We have seen more than one example of a sampler, on his own, deciding that the yellow stuff (sulphide) is better than the white stuff (quartz) and therefore taking a disproportionate amount of the former.)

Data collection also includes survey control. Use only one grid for the entire project and ensure proper checks are in place. There is no use cutting two metres of 30-gram material if you can’t find the sample location underground. Don’t minimize the importance of accurate survey control from the very earliest stages of exploration.

Do ensure the data are properly and consistently recorded and presented in a user-friendly form. Data that are too cumbersome and difficult to use will be disregarded. Equally useless are data that are stored only in someone else’s head.

Do make sufficient bulk density measurements, taking the ore porosity into account.

Do cut stray high assays if they cannot be properly explained, since most gold deposits are cursed with some degree of nugget effect. To paraphrase an old saying, cutting is a little like democracy — it’s a lousy system, but it works and what are the aternatives? Taking the axe to 5% of the assays quickly makes mineralized rock out of ore. If you decide not to cut, get one or more outside opinions on the matter.

Related to the practice of cutting is the cutoff grade. Most estimators choose a value based on the perceived economics of the project at the time of the estimate. When it comes to actual mining, the cutoff grade will be lower, unless stringent, efficient grade control programs are in effect. A grade control program that actually controls production is a rare occurrence indeed.

Do be realistic in assessing dilution, remembering our earlier example. Use a specific width of external dilution suitable to the conditions rather than an arbitrary fixed percentage applied to the final tonnage.

Do process a bulk sample under carefully controlled conditions to test the predicted ore reserve grade of a specific quantity of ore against its assayed head grade. This is an absolute must in every gold project, perhaps with the exception of a Hemlo-type deposit.

Use a sample tower, if not a pilot plant, to take statistically significant samples of the mined material. The bulk sample is also an excellent opportunity to confirm the metallurgical testwork.

Do conduct a stope mining test if there is concern over what dilution rate to use.

Do prepare a formal mine life plan and production schedule to develop and mine the minable reserves in an orderly, logical fashion. Mine the whole orebody out on paper, where mistakes can be made and corrected cheaply.

Do approach geostatistical reserve calculations with caution. Even the most ardent proponents of geostatistics will concede that their science (art?) encounters the most difficulty when applied to vein-type gold deposits. To compound the problem, this type of treatment is very poorly understood by the average mining engineer and geologist and, as a result, the output is frequently misapplied. Despite the best efforts of the geostatistical practitioners, geology is often forgotten. As a geologist, you can be sure that when you hand over your ore reserve block plan to a mining engineer to develop his plans, if the geology is not already taken into account, it never will be. And do remember that a geostatistician, if he were truly interested in geology, would perhaps still be a geologist. He is the computer jockey — you are the ore reserve estimator.

Do, in all areas of the development and production plan, build in flexibility to allow for things to go wrong. This means more reserves, more development faces, more stopes, and more pre-production development than your calculations show you need. Many of the mines we have studied have begun production with only a minimal amount of working capital, expecting to fund development from production.

Do be prepared to subject your plans to one or more critical reviews from disinterested parties. Listen to what they say.

Do be prepared to discuss your mistakes — we can all learn something from every experience.

Dont’s

Do not ignore or rationalize bad news or negative data. All data tell you something and must therefore be incorporated into your plans.

Do not get too hooked on computers. Although they are fast and provide much more data, they are not a substitute for judgment. Ore reserve estimation and mining itself generally are still very much arts and some parts of the decision-making process cannot be reduced to logic alone.

Do not listen to anyone who says “we are going back to get the high grade the old boys left behind.” The old boys didn’t leave high grade and they were a lot hungrier than we are.

There is an old saying that a pessimist is an optimist with experience. A series of over-optimistic decisions in the past decade have given us a lot of practical experience. Since no one wants to be called a pessimist, why don’t we strive through the 1990s to be realists?