Critical Elements Lithium Corp V.CRE

LI DMS is complicated

[goodtoreadthis]

DMS stands for Dense Media Separation if my memory serves me correctly. 

1. Guido, I don't agree. Anybody who has built a DMS, gets to learn very quickly, the limitations of individual pieces of equipment. As a start point, the main cyclone has to be sized to accommodate the expected feed rate. The feed gravel, typically between 1mm (don't believe anyone that says -1mm can be effectively treated in a DMS) and anything up to 50mm (50-100mm particles can be treated in a LARCODEMS unit) is mixed in a ratio 5:1 with circulating media ( a mix of water and finely divided ferrosilicon powder). If a 500mm diameter cyclone is chosen, it will only treat up to 100 tph of -32+1mm gravel. A 600mm diameter cyclone will treat approximately 150 tph of -40+1mm gravel. Once the cyclone is selected, the screens have to be sized to accommodate the flow of circulating media (to drain back to correct media sump) and a wash section to wash off all of the adhering ferrosilicon off both the floats and sinks products.
   
   There are typically four pumps in a simple, two stage DMS and all of those pumps must be accurately sized to perform their fixed duty. Likewise, the magnetic separator has to be sized to recover about 99% of the ferrosilicon which reports to it, sending the ferrosilicon back to the correct media sump and the effluent either to the feed preparation screen or to thickener, for recovery of clarified water. In my long career, I have never seen a DMS treat more than about 5% of its design throughput. There are so many pieces of equipment, which have to be matched in terms of performance, it would be an absolute MIRACLE if each piece of equipment ended up treating 170% more feed than it was designed for. It just doesn't happen. DMS plants often end up treating reduced feed rates (and we were normally very careful in listing possible causes for poor throughput) mainly because of flat material, very angular material or very friable material. When material is porous, ferrosilicon can enter the particle and change its density, lead to reduced product yields and lead to increased ferrosilicon losses. Ferrosilcon powder is not cheap! The last I remember it was costing about $1000 for a 250 kg drum. If an operator has a bad shift, he can lose 2 drums of ferrosilicon and lose 3 hours of downtime, chasing his tail troubleshooting blockages and other problems. He has to take numerous hourly samples, checking for holed screen panels, densities of different streams and checking automatic density control and water addition make up valves.
   
   When a DFS is performed, a reputable lab such as NAGROM or ALS perform extensive testwork on ACTUAL ore, so it is very possible to predict full scale plant performance based on this. Certainly, there can be surprises and hiccups. For AJM and PLS, their ores are pretty consistent and well understood. a large proportion of their ores have to be ground fine, because of the fined grained nature of the ore. The labs nowadays can perform lab testwork which can predict almost with 2% accuracy, the grinding energy required to reduce a sized product down to the grain size which most effectively liberates the optimum quantity and techniques to size tumbling mills are incredibly accurate. I would be most astonished to hear of a mill somewhere performing 170% above nameplate. It is unheard of in the business. If it were to happen, the supplier could almost be accused of supplying too large a unit! Experienced design teams check each other's work and supply fit- for- purpose mineral processing equipment.
   
   My feeling of JB and the BOD is that they are all straight shooters and I would be most surprised to hear of claims of absurd throughputs, made only to fool the unsuspecting investors. Conversely, I have confidence in the equipment suppliers and the design teams on this project and I would hope that any throughput deficiencies are only temporary in nature. A reputable design team like DRA, have about 40 years experience is discharging difficult, complex projects worldwide and I am quietly confident that they will achieve nameplate throughput relatively quickly, with a few months of ramp-up to get up to consistent feed

   Share

   AJM Price at posting: 22.0¢ Sentiment: Buy Disclosure: Held

   44 Upvote

   28Great analysis

    Report Reply

    
2.  
3.