Ann: Carbonatites Deliver Thick, Near Surface REE Results, page-65

"I see an economic mine in C3 already. There's a pretty obvious 30-50m deep open pit containing ~1%TREO + titanium + scandium ... literally mineralised from surface - basically zero strip (obviously depending on geotech and geomet)."... good to have an exploration geo's optimistic view, though 1% TREO is very low grade by industry standards and of course the metallurgy is critical. Yin is a great deposit, but not enough to carry a HAS style standalone development for reasons I've posted previously.

Let's dig a little deeper into C3's supergene 'orebody' as it's taking shape. Looks like a late stage fluid pathway system, centred around approx 400m x 400m thick, HG core (as defined by supergene oxide in any case). Most existing or next-development carbonatite mines (REE, Nb etc) are situated in the weathered supergene ore of lateralised carbonatites (certainly not all), where the downside of poorer recoveries is outweighed by the upside of grade enrichment though weathering deflation. NdPr dominates LREE carbs and is realistically 90% of the REO basket payability and rising. NdPr or Mine Reserves for comparison below. Yangi (outdated) demonstrates the advantage of high NdPr ratio despite low TREO % grades (~1%). Mountain Pass achieves 1% NdPr from 8% TREO, Mt Weld 2% NdPr from 9% TREO...



C3 at 1% TREO and 22% NdPr:TREO ratio = 0.22% NdPr which is very low reveune/t when each tonne is very expensive to beneficiate, and process with hydromet. REO-plants run about 1Mtpa +/-, so it';s grade, recovery, payability, recoveries not large tonnages that are required. Still, we can put an early guess on size given the robust style of supergene oxide mineralisation. 400m x 400m of high grade C3 core with an average 25m depth as per significant results to date = 10Mt @ 1.17% TREO, 0.26% Nb2O5, 4.7% TiO2



That table contains all C3 HG oxide intervals. Roughly 25m thick orebody, under 16m avg overburden for a 0.65:1 strip ignoring geotech, bunds, ramps etc. A rule of thumb in resources is that lower grades intersections and the reality of grade averaging mean a final MRE will have lower grades than the better holes, perhaps more tonnes, though in this example we assume a simple large block. Rarex has a similar HG core to their Cummins carbonatite (monazite dominant RE) where a lot of drilling has been done over many years. Cummins HG core is approx 50% larger than C3 as I outlined (at this early stage), with the 2021 MRE reporting the following at diff cut-offs (in addition Cummins MRE also has 135ppm Scandium avg).


When I say similar, Cummins has a high-grade core to the orebody that DRE has not tagged yet, if it exists at all. A type section from the MRE release demonstrates the HG breccia core, while variable along strike is present in similar grade and thickness (TREO width and % grade highlighted yelklow).


The 18.8Mt @ 1.15% TREO demonstrates what I mean about high grade holes invariably averaging down in the MRE estimation process with lower grade holes into a lower grade global tonnage. REE, or any of the deposits previous owners, have not been able to make this supergene deposit work because of low recoveries and RE-concentrates typical of supergene deposit. Now the plan is focussed on a phosphate development from their >30% P2O5 phosphate dykes, then hopefully leverage off the infrastructure etc to develop the REOs. Mt Weld is 8% TREO, Nuagulla is 4.8% TREO, Longonjo is 2.5% TREO ... metallurgy is critical but it certainly helps starting with high TREO grades...

I was hoping C3 infill drilling would tag much higher TREO supergene grades than first pass but it didn;t. If C3 (or elsewhere) doesn't have some serious amount of >2% TREO intersections at 22% NdPr ratio, then it simply puts DRE's carbs <1% MRE and possible mine reserve way back in the economics line. HAS is struggling to make economic sense at 0.35% NdPr and and 0.24% NdPr is their mining cut-off. Sure mining is cheaper in a large oxide pit, but mining cut-off grade is determined by the cost of beneficiating and Hydromet because 'mining' costs are considered sunk and what matters is margins through the plant vs onto the waste dump.

What about the Nb, Ti and Sc bi-product credits? Numbers look good in a spreadsheet, but this is where mineralogy and metallurgy get diabolically difficult and expensive. I admit to reading up on some Nb and Ti papers last night because I'm not a geology nerd and these carbonatites are complicated. Niobium is only mined in a few deposits globally, with the Araxa in Brazil producing 85% and Niobec in Canada 10% of global supply (in a small market mind you). Both carbonatite hosted pyrochlore deposits. The Morro dos Seis Lagos Nb (Ti, REE) deposit in Brazil (Brookite, Anatase, Rutile hosted mineralisation) is a massive 2.9 billion tons with 2.81 wt % Nb2O5 and can't even get off the bench to play...
Ti Nb mineralisation Carbonatites Russia 2018.pdf (excellent though very technical paper on Ti-Nb carbonatites).

C3 assay results to date look like classic Ti-Nb-REE carbonatite deposit, and the TiO2 grades are very high. As an ex min sand geo, these TiO2 numbers got me a little chubby, and I thought maybe titanium would be the money shot? As per the table above, outside hole 85 flyer with 10m @ 1.2% Ni2O5, C3 is much lower grade average with a sub-0.6% Nb2O5 (LHS axis) around 0.27% over thicker zones including hole 85. As per the scatter graph below, titanium is 10-20 times niobium... C3 is a titanium rich, niobium relatively poor carbonatite zone, and the Niobium grades look too low for a couple of reasons to be economic (from early days speculation of course).


PS... Hole CBR051 is a data entry error, cuplicating 1.9% P2O5 in the otherwise Niobium poor C5 carb


Why I say niobium is too low isn;t just because 0.27% N2O5 is an order of magnitude lower than some competitors, it;s because C3 is likely to be brookite-rutile hosted, not pyrochlore hosted. Pyrochlore has a bad rap for poor recoveries, where it breaks down into very fine grains during milling and 15-25% of the niobium is lost to slimes in the tails. But TiO2 (anatase-rutile-brooktite spectrum) can be difficult to recovery in both the ben plant and hydromet due to grain size. Globally, niobium is seen as very late stage over-growth and replacement of TiO2 minerals that were mobile and formed by earlier but still ate stage fluid metasomatism. The titanium minerals in the Russian carb study are similar to what I have seen in some many other carbs... multiple minerla species, very fine grained and intergrown with all sorts of other stuff (yellow scale is 200 micron). Don't believe there has ever been TiO2 recovered as a bi-product from a carbonatite mine...


I don't want to get technical when it's early days for the metallurgy. Gifford Ck ironstones are somewhat unique in their coarse monazite mineralisation with high NdPr ratios. It's what makes them potentially economic despite low grade/tonnes and high strip mining. Carbonatites are usually finer grained with multiple and problematic mineralogy for metallurgical recovery. DRE have done some mineralogical studies over the last 4 months obviusly, and made the following comments.

"REE mineralisation has been confirmed in both weathered and fresh carbonatites with petrographic workshowing coarse grained (up to 0.25mm) monazite and REE carbonate mineralisation in P-rich ferrocarbonatites and Ba and Ferich magnesiocarbonatite"... How much of C3's TREO is tied up in REE carbonates vs the Yangi style ironstones.. probably a lot, so cut whatever TREO grade C3 ends up with harder to comp with Yin. How fine is the REE-Nb-Ti minerals in C3 verse the coarse grained monazite of Yangi style ironstones... probably a lot, especially in oxidised ore where even the monazite can dissolve and remobilise.

In summary, a lot can go wrong with economically extracting REE-Nb-Ti etc in carbonatite hosted mineralisation, and it generally takes years and lots of work to resolve a workable solution. Therefore, I start by wanting to see high primary grades to indicate good probability for economic extraction despite expected recovery losses. C3 at 1% TREO and 22% NdPr just doesn;t do it for me, even if they leveraged off an Yin+ ironstone plant and infrastructure. Likewise the Nb and scandium is too low for my liking, given it requires separate plant to float, concentrate and process through an entirely different Hydromet plant. The Titanium grades are fantastic if the mineral size, recovery and chemistry is clean enough to sell into the pigment industry, but what I read says it probably won;t be. Time will tell, meanwhile the market will price risk-reward for C3 appropriately... which isn;t much imo.

GLTAH