Separation of Rare Earths – Art vs. Science (Part II)
In my last article I was anxiously awaiting my first site visit to a fully-fledged Chinese REO separation plant. Since my previous extensive experience with uranium separation solvent extraction plants and my exposure to REO processing at Mary Kathleen, I was very keen to see how technology had advanced, process design improved, control systems enhanced, particularly on separation circuits producing multiple product streams, exceptional quality, with modern plants. These first visits were to the JAMR and ZAMR plants that became part of Neo Materials and now Molycorp. What did I see? Well to be quite truthful – nothing. Yes, there was the expected 10’s and 100’s of stages of solvent extraction mixer settlers formed into banks of primary extraction and stripping, similar progressive systems for the less dominant REOs, very manual feed preparation and product finishing and handling , but where was the control system with it’s new age instrumentation supported by thorough process logic, detailed observation and monitoring? Where was it? I recall discussing this with Dudley Kingsnorth (also on the visit). The question was how can something (REO separation) that is stated as being so difficult to do, appear to be so easy? The process just ran! There were no control instruments. There were no variable speed drives. There was no readily apparent method of checking for phase continuity, no apparent ability to adjust phase ratios. The only process measurement I could see was a packet of litmus papers (used to measure pH), and from the readings chalked on a blackboard every hour there was not a great deal of variation occurring. I was gob-smacked! I was deflated! I had come here to learn but could see nothing to learn from. How can this separation that I have been told is so difficult, be so easy? So easy that it does not need the degree of control that I would have expected.
Standing in one of the world’s leading separation plants I had a Confucian moment! I had come here looking to take away answers! I had not come here armed with the right questions! So, I have spent the last few years working out the questions and hopefully developing the answers. I have also spent a lot of time developing friendships. Back in the mid-2000’s I was seen as a stranger, almost certainly a competitor, a little arrogant and certainly quite ignorant. Why should the Chinese REO separation gurus give up their secrets? I was determined to do whatever was necessary to not be seen as a stranger, to become a colleague in REO, to learn humility and to be open to learning. I now have many friends and colleagues in REO separation and we collaborate. I will share with you our philosophy and thoughts, but not just yet. First, you need to understand what the current term “separation” that is banded about in the investment literature is not.
Scientifically speaking, separation is just that. Separating one from another. In the REO scenario, again scientifically speaking it is the taking of a liquor based solution of a mixture of REO and impurities, and getting the various elements or compounds separate. Just imagine a cup of coffee. You want to treat this mixture to get the coffee back, the water back, the sugar and the milk (or cream!), all in nice clean separate piles. In REO, the process used is solvent extraction where a specialized and very specific organic compound (in liquid form) is used to target the elements of interest. However in the REO industry there are a number of solvent extraction steps that are not included in the accepted term REO “separation”. For example one, in the processing of REO in Niobium/Zirconium hosted ores, after baking and water leaching, a sulphuric acid based liquor is formed that contains the dissolved REE, Nb, Zr, co-extracted aluminium and many other impurities. The Nb/Zr is separated from the REE and impurities by solvent extraction, then the REE is separated from the impurities by solvent extraction. These solvent extraction stages are not part of the current term “separation” as they are part of the extraction circuit. For example two, in the processing of REO in conventional uranium circuits, a solvent extraction stage is used on the liquor tailings from the uranium circuit. Again, this is not part of the current term “separation”, as it is also part of the extraction circuit. And again, for example three, in the processing of REO from phosphate based fertilizer circuits, solvent extraction is used to concentrate the REO away from carried through phosphate. Again, an extraction circuit function. So what is the current term “separation”?
In almost every case I can think of, the final product from the REO extraction circuit is an REO compound in solid form. It is either a rare earth carbonate, oxalate or oxide and is a mixture of all of the individual rare earths that were present in the ore, and very importantly plus some impurities. There used to be a market for concentrated liquids but I am not aware of this practice any more. I have heard Jack Lifton reference the term here as PLS. This is an incorrect term here. PLS is pregnant leach solution and in reference to the current term “separation” is the liquor after the solid REO compound has been re-dissolved. So for overall coverage, the solid rare earths mixed carbonate or oxide is the feed to the separation plant. This is not to be confused with mineral concentrates, these are the feed to the extraction stage prior to the separation stage. So mining, mineral processing to produce a mineral concentrate, extraction to produce a mixed rare earths carbonate or oxide, then separation. Just as “oils aint oils”! Separation is not just separation. A separation circuit for LREO is different to a MREO circuit and is different to a HREO circuit. The circuit that you design (or visit) has been specifically built based on the spectrum of REO in the feed, the range of impurities in that feed, the products that are desired to be produced and very importantly the quality of each product to be produced.
When I first got seriously involved in the REO space, I looked at research done on the value adding chain of REO and the companies involved. Obviously China was pre-dominant in the early part of the value add chain. I was amazed at the vast number of quite small operators there was. There were many producers of neodymium for example, each producing a few tonnes, sold to the next in line and onto the final end user. The point that I finally realized here was that the reason the supply chain looked so wide and dis-jointed was because it needed to be, because each end user had a different quality specification. So what this means to a separation plant is that you don’t simply have a market for cerium oxide, for example. There are many products called cerium oxide, and they are all different. Different depending on where the product is going and what it is used for. Different quality. Different impurities. Maybe different size distribution. Wait for it – certainly different packaging. If we follow the cerium oxide example, if it is used for glass polishing, or as part of a catalyst recipe, or as an additive to automotive glass (Lexus), it will have a different specification. Where am I going with? Well, as I try to take you on the separation journey, you need to understand that the design and operation of a separation plant is as much about where the products are going, as to where the feed stock came from. So there is no one design suits all, no generic design. Each design is a very specific statement about the feed material and the products to be produced and their destination. Next week, I will relate a few examples to give you some insight into what types of plants are out there and what and why there are differences.
Get our daily investorintel update
Mr Mackowski is a qualified engineer in mineral processing with over 30 years technical and operational experience in rare earths, uranium, industrial minerals, nickel, kaolin ... <Read more about Steve Mackowski>