Geology & GeoMetallurgy

Geology & GeoMetallurgy

  • To participate in the 911Metallurgist Forums, be sure to JOINLOGIN
  • Use Add New Topic to ask a New Question/Discussion about Geology and Geometallurgy.
  • OR Select a Topic that Interests you.
  • Use Add Reply = to Reply/Participate in a Topic/Discussion (most frequent).
    Using Add Reply allows you to Attach Images or PDF files and provide a more complete input.
  • Use Add Comment = to comment on someone else’s Reply in an already active Topic/Discussion.

What is the Conversion Rate of Resources & Reserves into Metal (16 replies and 3 comments)

(unknown)
8 years ago
(unknown) 8 years ago

Investors and Policy makers place a lot of faith in the assumption that (hopefully) all of the company's reported reserves and (most of the) resources will be mined. The key question is, empirically what are typical conversion rates achieved by industry? What is the Conversion Rate of Resources & Reserves into Metal?

A recently completed a high-level study looking at the factors behind the recent decline in the Reserve & Resources of some of the major gold mining companies. The process came across a really great set of (public) data on AngloGold Ashanti - titled "Mineral Resource and Ore Reserve Report" (which you can download from their website). Each year they generate a very comprehensive set of notes on their individual mines and projects. It also includes a table which details the factors behind the change in the reported Reserves & Resources. Over the last 5 years their MI&I Resources didn’t change much (from 227 Moz at the start of 2010 to 231 Moz as at the end of 2014) but “under the hood” lots of things happened. They discovered 48 Moz of gold, but depleted their resource base by 29 Moz (through mining of 21 Moz of gold); they also acquired 4 Moz from other companies and divested 11 Moz to others. The rise in gold price created 14 Moz of extra resources (through using lower cut-off grades) but rising costs took 12 Moz away. Other factors, including changes in modelling assumptions and scope of work (such as shift from open pit to underground) reduced the Resource figure by 4 Moz.

Over the 11 year period (2004 to 2014) AngloGold Ashanti produced 54 Moz of gold but, in the process, they depleted 64 Moz of Reserves and 75 Moz of Resources. In other words, on average they were only able to convert (54/64=) 84% of the P&P Reserves and (54/75 =) 72% of the reported MI&I Resources into gold bars for sale. The implied loss factor of (100/84 =) 19% for Reserves and (100/72 =) 39% for Resources is a lot higher than the headline metallurgical loss factor of 8-12% through a conventional mill. This suggests that other factors are at play (such as economics which leads to large parts of the orebody not being mined). Given that so much of our resource base will never be mined, this suggests that out future available of metal is much “tighter” than you think – with obvious implications on the future commodity price.

The above analysis is at the company level. AGA provide these numbers on an annual basis for each operation. It would be an interesting study for an Honours Student to unpack and analyse the variability of the data at the mine site level.

A couple of questions:

  1. What’s your experience on the reserve and resource conversion rates at your operations? ie are AGA’s numbers representative of the industry?
  2. Do you know of any other companies that publish a reconciliation report like that published by AngloGold Ashanti? Am looking for a broader sense of conversion / depletion rates for other companies and commodities. 
Helena Russell
8 years ago
Helena Russell 8 years ago

I assume gold as the leading point. Highly variable. The Geological Survey of India and the Indian HUTTI mine of Raichur District of Karnataka State, ran statistics on a single Strike Reef. The result was around 63% of resource estimated recovered in the underground section. There was no open cast there to compare. A geologist who's worked there many years tells me the Main Reef, mining of 10 years, gave barely 34% recovery from the estimated resource.

(unknown)
8 years ago
(unknown) 8 years ago

This question can be dangerous because people will take a ratio from what I assume is a reasonably documented deposit/geology/mining/milling costs and efficiencies and try to apply it elsewhere. Here is my favorite example. There is a deposit with a resource of +10 million ounces of gold. Note the word 'resource.' The thing is that it is a huge mass of about 1 gram gold in mountains adjoining a river. How many people really think that many hundreds of millions of tons will be permitted, mined and put on dumps and leach pads in mountainous terrain right next to a river? What will be the resource conversion rate to reserves, let alone metal, there?

Helena Russell
8 years ago
Helena Russell 8 years ago

That is an interesting quest but the unquestionbility of piecemeal data such as mine, furnished above, may not lead to any tacit generalisation. As the question came up in the discussion, I thought, my information could be some supplemental, for further precise thinking.

JohnnyD
8 years ago
JohnnyD 8 years ago

The issue under debate it is more complicated than it can be commented in a few words.
Firstly there are different types of gold deposits that were estimated as reserves. In this warehouse the gold may be presented differently in terms of Au mineralization themselves. According to this it will differ the extraction costs.

Secondly the deposits may have different morphological configurations: veins, stratabound, skarn, IOCG, orogenic-Au, etc
Each of this type will have a specific operating methods that are chosen according to certain criteria:

  • which would be the loss reserve in operating/exploitation
  • which would be the loss in the technological method
  • especially in underground
  • which would be the lost in extracting ore from hewing

I think that you can compare it if:

  • the deposits are the same type;
  • the gold-bearing mineralization has the same features (we can not compare a Carlin-type ore with Withwaterstrand ore or a IOCG gold type ore);
  • the reserves are almost identical and have same exploited average content (of deposit);
  • almost identical methods of operation and ~ similar loss of metal and reserves; etc. etc.

From this few presented characteristics would not get any benefit a reserve can turn easily in resource.

Bob Mathias
8 years ago
Bob Mathias 8 years ago

If one really want to make the reserve data meaningful, it is always good to dig into the metallurgical data. Normal quarterly company results show how much are mined and the grade. Some companies provide recovery data but it not available, it is easy to work back on the raw data and get the efficiency of the mining and metallurgical departments. Looking into the cost per lbs or cost per ounce are also a mirror of the efficiency or difficulty of mining and processing.

(unknown)
8 years ago
(unknown) 8 years ago

I think that Bill made a good point by showing on his example that resources and reserves may overestimate quite a bit the metal that can really be extracted. And the estimates in question were done by the company that had all the means and interest to do it properly. Which, in combination with my personal experience, suggests that this could be a general trend, notwithstanding the mineralization type. There could be different reasons for such discrepancy. Classifying inventory as resource, as suggested by Tony, is one of them. But there should be other reasons as well, and it would be good to identify them.

(unknown)
8 years ago

I think that management in many cases has a big incentive to please the market and they hire consultants that (consciously or not) would be willing to please them. The estimate might look very reasonable, but not entirely feasible.

Helena Russell
8 years ago
Helena Russell 8 years ago

Aware of this, the Indian Bureau of Mines, with which I had good association in past even for discussions and consultations, adopted a term "Recoverable Reserve" in the IBM Year Books for the last decade or even more.

The term 'recoverable reserve' is unnecessary, confusing, and very poor practice. By definition, a reserve is both legally and economically mineable. For a feasibility study, a scheduled ultimate pit is required for a reserve. A non-recoverable reserve would simply be a resource.

(unknown)
8 years ago
(unknown) 8 years ago

Good point, according to the CRIRSCO Template (JORC, SAMREC etc.), when announcing to the public, only Proved and Probable Ore Reserves or Mineral Reserves can be stated---"recoverable reserves" not a good idea. However on the other hand Ore Reserves are a constantly moving feast as some of the Modifying Factors such as commodity prices are constantly changing. Great discussion, well done to all and thanks Mr. Fraser for initiating it ----interesting to see if there is any further quality information available.

(unknown)
8 years ago
(unknown) 8 years ago

A geological model is always an oversimplification of the reality. If not properly addressed, this more likely than not results in overestimation of the resource/reserve, particularly in areas with complex geology and/or high grade variability. Normally, estimation errors should also be reported. But they are difficult to assess and JORC allows for qualitative discussion of the uncertainties.

I agree that the term "recoverable reserves" is a bit confusing and should better be avoided, but it is in common use in many countries since quite a while. Ward (1984) refers to them as "the amount of coal that is actually expected to be won from the ground, using technology prevailing at the time of assessment..." as opposed to "in situ reserves".

Ward proposed a very simple factoring of recoverability: - 10% for open pit and - 40% for underground mining. These days this should probably be accommodated by resource/reserve estimates.

I guess Ward based his rule on the statistics available at the time. If we forget about coal and apply the Ward's rule to the Bill's analysis, assuming open pit operation, we will get a much smaller discrepancy between the estimated and recovered metal. That is probably Bill's interest, to gather statistics on this kind of discrepancy for gold.

(unknown)
8 years ago
(unknown) 8 years ago

I do not understand your factor comment. Are these ore losses to be applied to a resource? A reserve would have a pit design that includes both ore loss and waste rock dilution. For example, an iron ore open pit might have 6% dilution and 5% ore loss. The more complex the deposit, the larger the factors.

(unknown)
8 years ago
(unknown) 8 years ago

A reserve estimate should normally accommodate ore losses and dilution. And I think that is what AngloGold Ashanti have done in their estimates. Still they have ended with a discrepancy of 16%. 

As I have mentioned earlier, any model is an oversimplification of the reality and this is very likely to lead to overestimation. This could be a possible explanation. Statistics from other mines would help to better understand the problem and may be find a factor of maximum deviation. 

JohnnyD
8 years ago
JohnnyD 8 years ago

In addition to all said above:

A GEOLOGICAL RESERVE identified and assigned to different categories by exploration works (drilling, galleries, exploration wells, etc) becomes INDUSTRIAL RESERVE when executing specific preparation works for exploitation (in operating block, panel, etc) adapted to specific method specially in underground (for eg,understorey gallery, etc).
As I said in my first discussion the recovery of the metal from INDUSTRIAL RESERVE depends on the type and characteristics of mineralization (dispersion, degree of homogeneity of ore, mineral gran size, etc).

Therefore: GEOLOGICAL RESERVES minus LOSSES => INDUSTRIAL ORE
The LOSSES can be: TECHNOLOGICAL LOSSES (specific to the method of exploitation; for eg.: loss in pillars + new sterile areas or under economic areas, identified in the exploitation processes; for eg: tectonic lamination of a lens or shear-fault along the vein, etc).

All these diminishes the industrial reserve (as quantity and content/metal).
INDUSTRIAL RESERVE (ORE ) - LOSS OF TRANSPORT => SUPPLYING/FEED ORE (at flotation)
SUPPLYING ORE X EXTRACTION YIELDS => CONCENTRATE METAL (M)

By a simple subtraction:
geologic (M) - concentrate (M) => Conversion Rate ; (calculating percentage [%] ).
All this is regulated/adjusted periodically by MOVING RESERVES [quantity(+/-), content (diff.), and metal (+/-)

(unknown)
8 years ago
(unknown) 8 years ago

Another factor to consider is loss of grade, not just tonnes. A massive study was done on mines that were exhausted during the 20th century (unfortunately I no longer have the reference) that showed over 90% produced more tonnes at a lower grade than their reserve estimates. Over digging and higher than projected waste rock dilution is quite common in mining. In addition, process plants are designed to work within a narrow range of grade variability. When ROM feed is outside that grade range, (higher or lower) plant efficiency decreases and metal losses increase.

Helena Russell
8 years ago

Loss of grade, within the connotation of not getting the estimated grade, comes from association of gangues with the ores, the quantity which was expected to be easily separated out and the practical difficulties in trying to separate gangues as per expectations during the exploration times. In fact producing more tonnes with less grades is the exact result, as you've mentioned. If plant efficiency demands the optimum can be blended after productionand before feeding into the plant, in stead of feeding the ROM directly. Grade and tonnage vary inversely.

Victor Bergman
8 years ago
Victor Bergman 8 years ago

The real recovery of metal/non metal and the estimation of reserves and/or resources are controlled by a number of factors.

  1. The origin and distribution of metal in the deposit. genesis of ore.
  2. Structural controls of the deposit.
  3. Metal association with other metals and gangue minerals and grain size.
  4. The type of exploration method and various controls used for exploration, sampling method and analytical procedures.
  5. Methods of reserve calculation, bulk density calculation and other factors used for reserve estimation.
  6. Mining and metallurgical methods used. And a number of other factors (the list could be quite long).

With so many factors affecting the ultimate recovery it is natural that the ultimate metal recovery will be deficient by variable percentages.

We already know the difference between the reserves and resources, and, depending upon the stages of exploration for a company handling a number of operations simultaneous.

Simultaneously, constant progressive revision of estimates is a natural process.
Hence, calculation of final estimates from proved reserves will vary from metal to metal, geological environment to another geological environment.

However it should be possible to arrive at a certain range of 'loss' to be expected from a specific deposit. I feel manual calculation of reserves with 'personal hunch' may bring us much closer to both the estimates rather than through computer programmes.

Helena Russell
8 years ago

My old colleague, too has duly explained the various reasons which could account for grade changes of the exploited ore deposit at the production stage but as a thumb rule, grade and tonnage vary inversely with each other, total metal content remaining more or less unaltered.

(unknown)
8 years ago
(unknown) 8 years ago

Tthe fact that total metal content of the ROM doesn't change with dilution does not mean you will not lose product reserves. A common problem with resource reports is that they very good at estimating crude ore tonnes, but much poorer at predicting total metal recovery. Reasons include insufficient metallurgical testing, bench tests that do not fully replicate concentrator operations, better recovery on the bench than in-plant, and inability of the mine plan to consistently deliver the plant feed blend at the required specs resulting in product losses. It is product recovered, not product in-situ, that matters.

Helena Russell
8 years ago
Helena Russell 8 years ago

Total metal idea was a broad coverage of the theoretical metallisation potential. There is no cover for intransient testing or insufficient testing there in.

Please join and login to participate and leave a comment.