Originally posted by Oreb
The original CAPEX figure was based off a BFS-level design so not exactly a guesstimate.
Not sure what you mean by Stage 1 / Stage 2 unless you are talking about future plants? If so, other than the design and proof of concept work no I don't believe the cost will reduce significantly for a second plant unless once the first plant is up and running they identify areas where it has been over engineered. Another possibility is that once they have cash-flow from the first plant they go a different route to funding a second plant and don't need the level of gold-plating that was required to get the ECA finance across the line.
No byproduct opportunities that i'm aware of. The waste/reject material from the process is benign and suitable for use as fill/road base or something like that from memory.
Other than SMS' involvement being fairly critical to receiving the ECA finance offer, the guarantees from SMS for through-put and quality are the main benefits to the SMS design (and I believe these guaratees were required to receive the ECA offer). They've obviously 'bullet-proofed' the design to be able to give those guarantees, hence the CAPEX blow-out. In an operation like this where the product is targeting 99.99% purity, commissioning and ramp-up is often an issue but this risk is essentially covered by having SMS involved in this way. Don't forget that Iggy has stated that due to the very conservative assumptions and likely over-engineering to satisfy the financiers, there is a real possibility the plant will produce significantly more than nameplate - 6000t has been mentioned.
I believe the key to high profitability is the proportion of 4N - 99.99% HPA they produce. At present they have it ramping up to maximum proportion over 2 + years. Before they contemplate building a second plant they will work out exactly how to achieve that purity. Assuming the demand for that HPA is there it makes profitability of a second one quite likely although a study re- competitors/ competing technologies would need to be done.
(20min delay)