DETRIOT () -- In reviewing the English language articles, which discuss the supply and demand of tellurium for manufacturing cadmium telluride photovoltaic solar cells, I noticed a persistent error in the interpretation of U.S. Geological Survey data on tellurium fundamentals and usages, which seems to now pervade all such discussions. I need to emphasize that those who utilize USGS data are reading too little into the numbers, rather than too much.
Based on the discussion following I now believe that there may be enough annual production of tellurium to build and maintain a solar electric power industry, which can become an economically important component of the infrastructure, total supply and accessibility of electric power generation on a stand-alone basis; I now believe that there is much more tellurium produced each year than I thought or than is commonly reported. The difference increases the credible output of tellurium by a factor of between two and four, but the global total annual production may still be less than 500 tonnes a year.
I wonder if the run up, over the last few years, in tellurium prices to today's $100 a pound for 99.7% tellurium isn't more due to speculation and, perhaps, hoarding (i.e., advance inventory accumulation-than to a rise in demand as dramatic as the price increase would indicate).
Tellurium is produced mostly, 80%+, as a byproduct of copper smelting and refining, the balance is produced as a byproduct of lead and precious metals, gold, silver, platinum and palladium refining. To the best of my knowledge there is today no primary tellurium mine in operation. I do know however - and it is listed and highlighted in a report I will link to in a few sentences - that there is a site in China, which is considered to be a candidate to be a primary tellurium mine, but I can find nothing about the particular minerals or the extent of the ore body.
To see how the tellurium supply data got confused let's go back to the very beginning of investor interest in tellurium at the end of the last century, those many, 9 to be exact, years ago.
In 1999, the National Renewable Energy Laboratory (NREL), a taxpayer funded operation to carry out scientific studies, commissioned a study entitled "Assessment of Critical Thin Film Resources Tellurium." Note that the word 'critical' in government-speak has come to mean something without which a device or a process cannot be built or done. The NREL had determined that in order for the government to consider doing research on thin film cadmium telluride photovoltaic solar cells as possible alternatives to fossil fuel burning to produce electricity it must first be determined whether or not there was a sufficient, reliable and consistent supply of the critical raw material(s); in this case the material of which the supply base was to be examined was tellurium.
A private company, World industrial Minerals, was chosen by the NREL to do the study, and the result was published, in 1999, and is freely available online here.
The 1999 report is comprehensive. It lists in significant, but popularly accessible, detail the mines, the metals and the recovery processes for tellurium globally and annually. It states, unequivocally, and with mining engineering references, that "the majority of tellurium produced comes from the ...anode slimes that accumulate during the electrolytic refining of copper...."
The report then lists 30 companies around the globe that mine or refine tellurium to produce one of three grades:
- Tellurium dioxides-sludges and precipitates of variable purity, around 35%;
- Commercial grade tellurium: 99.7% purity;
- High purity tellurium: 99.9 to 99.9999%.
The report's author notes that, in 1999, because of [then] current low demand, several of the facilities listed no longer recovered tellurium. The author then points out, "If improving market conditions would warrant, these facilities could again recover tellurium with a minimum amount of effort."
The report then states that the global average of contained tellurium in one tonne of refined primary copper, is 0.4 pounds, and in one tonne of refined, new, lead is 0.1 pounds. The author then states his belief that, although recovery rates, in 1999, are typically, at best, 39% this recovery rate could increase to 50%, or more, if the demand were there.
Now look at the following tables and the conclusion following them from the report and pay careful attention to the amounts of tellurium "possible to be produced" annually in the U.S. and Chile:
"The combined production potential from copper and lead refining totals 859 metric tons of tellurium for 1997. The actual production of tellurium for 1997 ranges between 200 and 300 metric tons (250 metric tons average). The capacity utilization is estimated to be 29%. It has been indicated that tellurium producers are capable of achieving higher recovery rates if the demand was present. Because of the proprietary nature of the various tellurium recovery processes it is not possible to determine the ultimate recovery efficiencies. With current refining practices for copper and lead coupled with generally consistent yearly metal product output, it is estimated that 859 metric tons to 1,716 metric tonnes of tellurium could be produced per year if demand warranted."
In 1999, the global production of copper was just over 9 million tonnes; the report's author estimates that the possible global production of tellurium just from copper, and with a maximum recovery of 50% of the tellurium contained is nearly 800 tonnes per annum.
In 2007, the global production of copper was nearly 16 million tonnes. Thus the possible production of 50% of the contained tellurium would be 800 x 16/9 = 1,416 tonnes! Why is this number so far removed from the 2008 USGS estimate of global tellurium production for 2007 as indicated in the table below, and what are we to make of the remarks that follow the table?
Serious questions arise from the discrepancies which themselves arise from comparing the numbers from the 2008 USGS data with that privately gathered, at the behest of another U.S. government entity, the NREL, and published in 1999.
But first you need to be aware of the fact that the USGS numbers for world totals above are frequently used by those discussing the total supply picture for tellurium.
Here for example are two excerpts from a couple of very recent reports. First of all, incredibly, is Slide 16 from a Power Point presentation entitled "Materials Demand for the Rapidly Expanding Solar Electricity Industry," published on the Internet in March 2007 by the NREL:
Note for yourselves the global totals for "Refined 2006 MT" of tellurium and the statement of the reserves, and compare those figures with those published by the USGS shown above, in 2008, and with the figures shown by the earlier NREL report published in 1999.
Now here's some text from a 2006 book, which is a study of the "Future State of the PV Industry-Trends and Technologies":
Note that it is once again, supposedly, USGS data used to support the supply conclusions. But these conclusions, the ones above, and any of them that are based on the current common interpretation of USGS data may be completely wrong! The explanation for all this misinterpretation is really simple:
Look at the USGS chart above. Note well that it gives data for only 3 of the 18 countries listed in the 1999 NREL report as having production capacity for tellurium from their copper mining. Note also that only a single one, the smallest, Canada, of the top three potential producers identified in 1999, is recorded as having produced any tellurium, and that only two other producers of the potential 15 remaining are identified as actively producing tellurium. The second largest potential producer, the U.S., is shown with no identified production, because the USGS has a policy of not revealing competitively disadvantageous information in return for obtaining the data from American private producers.
In this case since there are very few tellurium producers published data on their output total would enable them to have information about each other, which they do not wish to share. The USGS always identifies such non-reporting with the symbol "W" as it does here. Finally you must note that the total production from 15 of the 18 potential global tellurium producers, including the two largest, Chile and the U.S., which between them had, according to the 1999 NREL report, a total potential production of more than 700 tonnes of tellurium as shown reporting either as W, explained above, or as NA, not available.
I have to say at this point that we must take into account when determining the annual global output of tellurium the fact that some copper production now uses a process for refining, solvent extraction leaching, that eliminates totally the recovery of tellurium from copper concentrate so obtained, and that such contained tellurium once lost is lost permanently. In addition, it may well be that recovery of 50% of contained tellurium is on the high side, but this must be countered with the fact that Canada with a potential, in 1999, for producing tellurium contained in 640,000 tonnes of copper managed to produce, even then, 59 tonnes or 50% of the total potential. Canada actually produced less copper in 2007 (507,000 tonnes) than it did in 1999, but it produced more tellurium, 75 tonnes in 2007 than it did in 1999 with 59 tonnes!
U.S. production of copper was also down in 2007 when compared with 1999, but Chilean production of copper in 2007 was more than double what it was in 1999! Finally, Peruvian copper production in 2007 was two and half times what it was in 1999, and its reported tellurium production increased, but no where near the increase that would have been expected were the 1999 figure, in the 1999 NREL report above, of a potential 17 tonnes at 50% of total potential been correct, even if Peru's recovery capability were only the traditional maximum of 39% stated in the 1999 article.
My conclusion is that even if we just reduce the 2007 potential 50% production of tellurium contained in copper by half and estimate that only 25% of the potential supply is recovered we then must recognize that the world today is most likely producing at least 500 tonnes per annum of tellurium or four times the commonly stated figure in supply reports!
Before the cadmium telluride PV solar cell enthusiasts begin rejoicing, I must ask some more questions. Even if there is enough tellurium produced to meet the demands of a growing cadmium telluride thin film using segment of the solar electricity industry, isn't there still a major bottleneck? Isn't there a serious dearth of capacity to produce the electronic grade tellurium needed by the PV solar cell manufacturing industry, and doesn't the building of new capacity to produce electronic grade tellurium depend on the availability of skills, equipment and the time it will take to get the purification capacity to match the production capacity hoped for? In addition isn't it time now for everyone to take a breather and ask the very important question: Are the numbers I have estimated and calculated here today correct?
One set of questions for the investment community. If the demand for tellurium is not immediate, and if the supply, is much more than has been stated, so that the supply exceeds the current demand, then why has the price of tellurium gone up so fast and so far? Isn't it time for a serious correction, but, if so, where will the investment come from for increasing the capacity to purify tellurium to electronic/solar grade? How much speculation is a necessary component of the total price of tellurium to drive the creation of increased supply and purification capacity?
I admit that in my previous articles on tellurium I misread the USGS data and followed the herd. Now I am either out in front, or I'm a Judas goat. In any case the tellurium supply situation needs to be addressed immediately before any more commitments to cadmium telluride projects are made.