Saturday, April 13, 2019

Pretium - Lies, damn lies and Reserves

I've been bashing Brucejack frequently over the 'difference' been mine grade and reserve grade, and with the release of the update resource and resource statement, I decided to look at how the reserves and resources have changed overtime (just for the VoK):

Note: I've added dashed lines to the charts where data is missing (the Dec 2016 resource press release didn't include information for M&I and inferred resources).

Tonnes




We can see that that we had an initial increase in resources from 2011 to 2013 and then a leveling off as drilling focused on improving resource quality and converting resources to reserves. We see that recently drilling has also focused on increasing resources.


Contained Ounces




Ounces have stayed relatively flt over time, we had very large growth in resources from 2011 to mid-2012, then the focus has been on improving the quality of the resources (from Inferred to Measured and indicated) and to reserves in 2016.

Average Grade



I've added a lines for the Life of mine (LOM) mine grades in the 2014 Feasibility Study (FS - orange line) and the new 2019 proposed mine grade (purple line), and one for the 2018 actual mine head grade (grey dashed line).

  • Average grade (all resources) decreased from 22 g/t Au (2012) to 15.8 g/t Au (2019)
  • M&I resources remained constant since 2012 at ~17.1 g/t Au
  • Inferred resources decreased from 35 g/t Au (2012) to 17.7 g/t Au (2019)
  • Reserve grade decreased from 16.1 g/t Au to 13.8 g/t Au
Obviously the reserve and resource grades have dropped because of all the high-grade mineralization mined in 2018......


You will see that in many 43-101 reports the Inferred resources typically higher grade than the other resource categories, they are normally based on the lowest quality data (relatively speaking), the widest spaced drill-holes and can be heavily skewed by a few ultra high-grade holes. However, it is surprising that the grade of the inferred resources have fallen by such a large amount.

This figure was also interesting in the press release:



There was a lot of technical crap in the section accompanying this table, but in I think it means they mined 20% more material than planned, now this will be interesting to see going forward, especially with the mine being expanded to 3800 tonnes per day.

Brucejack produces a lot of gold, but it keeps falling short.

Data for these charts came from:

Feb 2011 resources - link

Nov 2011 resources - link

Apr 2012 resources - link

Sept 2012 resources - link

Nov 2012 resources - link

Dec 2013 resources - link

July 2016 resources - link

Dec 2016 reserves - link

Apr 2019 reserves and resources  - link

14 comments:

  1. Basically they are comparing the volume excavated against the 2019 reserve model. The result shows 20% of excavation to be outside the reserve model boundaries. Hence 20% diluted waste. What they did not show is reconciliation against the 2014 model. But they did say that they would switch to londitudinal stopes from transverse, which implies they have high grade ore trends with waste in between that was taken in transverse stopes but will be left in place with longitudinal stopes. The new stopes will require more ore development but less in waste. This should help costs somewhat and also help with HW dilution

    ReplyDelete
    Replies
    1. Basically they were saying, out reserve model was excellent, we just didn't mine it properly and everything henceforth will be perfect...

      Now the big issues could be - can they mine and develop sufficient headings to feed a mill that is 30% hungrier with juicy rocks?

      Delete
  2. I agree, Table 4 shows actual 2018 mined areas vs. the new 2019 resource model. They are saying that had the stopes mined in 2018 been based on the new model, they could have mined more selectively, and achieved the reserve grade. It would be really interesting to see the shapes they based these comparisons on.

    They have included a lot of infill drilling in their new estimates, which probably helps. Key will be to show increased produced grades as a result of this in the next quarter.

    A question is how the drill density (risk) for reserves planned two or more years out, compares to the density of the data in the Table 4 comparison.

    ReplyDelete
    Replies
    1. However, they will always have the production pressure problem.

      2018 actual throughput = 2756 tonnes per day
      2019 reserve model, if mined = 2194 tonnes per day

      or only 79% mill utilization.

      This leads us to an interesting question - to maintain the reserve head grade with a mill that is 1000 tonnes per day (or 35%) bigger, can they be selective enough with their mining to mine enough material to feed the mill and maintain the head grade?

      Delete
    2. At 2,900 tpd the reserve base is large enough but at their proposed higher throughput rate the pressure will be greater

      Delete
    3. Yes, but with the more selective mining method (or a better control on mineralization), how many more headings (and therefore people and equipment) will they need to feed a 3,800 tpd mill?

      Delete
  3. No doubt there were excavated volumes outside the 2019 reserve outlines but what is perhaps more interesting is if there are reserve outlines outside the excavations. This would be ore loss from an inaccurate model.

    ReplyDelete
    Replies
    1. I'm guessing it was a face saving study to provide confidence in the reserve/resource model. It is easier to blame the mining method (it would just knock-off a few % of the share-price), but acknowledging a potentially fundamental problem with the reserve/resource calculation would be a massive kick in the nuts.

      Delete
    2. They had a serious reconciliation problem that was likely the reason for a new resource model...not uncommon for new mines.

      Delete
  4. They moved the goal posts. Should have compared actual mining to the previous reserve model not the new one. As I've said here and elsewhere, the VOK mineralization is continuous in corridors along the edges of the stockworks and at structural contacts. It's not individual veins, it's swarms of veinlets carrying visible gold that can be followed in the headings (if oriented on strike, which is mainly orthogonal to the transverse stopes). Bulk mining as if the high grade is like raisins in a loaf of bread was always going to be insane. Now it remains to be seen if they will swallow their pride and hubris and go ahead and mine this thing properly, or keep up the charade of a bulk tonnage mine (wherein they will continue to put lots of waste into the mill merely to keep it at capacity in which case they will have a hard time moving off the 10 g/t).

    ReplyDelete
    Replies
    1. Yes the goal posts were moved. It’s interesting they gave the 20% number for diluted tonnage, but total reserve tonnage did not drop from 2014 report to 2019.

      A proper reconciliation would show comparisons of both Tonnes and grade, with resulting ounces. They should also have two reserve models. One for LOM and one for short term (1 to 12 mo) planning. It would incorporate both exploration and definition drilling, along with actual results from nearby past mining.

      Delete
    2. I should have said “did not drop that much”

      Delete
    3. The overall reserve grade dropped by ~17%. What is interesting is that the proven reserves dropped to 11.2 g/t Au, which is very close to the planned mine grade for 2019 (10.6 g/t Au).

      So, lets see how crap Q1 2019 will be...

      Delete
  5. Hello Angry Geo could you do a write up on MXR press release Friday after 10 weeks of waiting on promised assays from Feb 14 2019

    Thank You

    ReplyDelete