Hi TraderK, I'm not losing sleep over it. To be honest, I have...

Hi TraderK, I'm not losing sleep over it. To be honest, I have not entirely figured out how they are managing the ponds but the conference call Q&A filled in some of the gaps for me.

I am confident that the ponds are big enough in total and was reassured to hear that evaporation rates are running higher than design rates (there were some safety margins applied when they scaled up from the pilot plant data). No problem there.

They target a concentration of 0.7% Lithium in the harvest ponds - at lower grades the output of the primary circuit will decrease, more than this and they start losing Lithium to precipitation of schoenite (Li2SO4.K2SO4). So, this target grade is very important in terms of output and yield.

But how to get from feed brine to concentrated brine... There are many ponds involved each with their own chemistry. It is important to realise that the ponds do not get filled - they have to actively manage the grades and inventory at each stage. At each stage/pond they evaporate water, they precipitate various salts depending on concentration, pH and temperature , and they will lose a little brine liquid that is entrained in the precipitated salts. The resultant grade depends on all of those things plus residence time (which in turn relates to flow rates relative to the volume contained in each pond). So to manage the grade they need to have made all the correct assumptions about evaporation rate, precipitation and entrainment rates and then manage the volume(inventory) in each pond appropriate to the planned production throughput at the end of the pond system. It is complex and so they have a very large spreadsheet to do this mass balance for them.

All good so far but it appears they had a formula error somewhere in this spreadsheet that was causing them to gradually accumulate volume at the front end and slowly starving the harvest ponds. In addition, it sounds like they might be suffering a slightly higher brine entrainment rate in the sylvite that precipitates in the harvest ponds - not a big deal as they will eventually remove the sylvite and drain it - just not yet... And one more thing to screw things up a little - if you are taking measurements of volume, you can't just wander out and check the surface level you have to know where the bottom is - that is easier said than done when precipitates are accumulating unevenly across the bottom of the pond.

So yes, a few things have come in to play to short change the harvest ponds of inventory whilst leaving a little too much in the early ponds. They seem to have got it figured now - just a few months of flat production while they pull volume forward in the process. It looks like the lesson they only need to learn once.

For what it is worth, as much as a certain commentator keeps banging on about how good SQM is, the GM at Olaroz, Cristian Saavedra is an ex SQM production manager - 9 years there - definitely started kicking goals when he came on board. Further back, the process was designed by Peter Ehren who had served as SQM's R&D Manager - also 9 years there. The team at SDJ are not lost in the wilderness as someone likes to suggest. Just a matter of time...