A more rigorous method would be to use estimated number of EVs x estimated weight of lithium per EV battery then convert to LCE basis, but to avoid the complexity of trying to estimate an average weight per EV battery, which depends on vehicle size, range, efficiencies etc, for now I just used the info here https://www.dakotaminerals.com.au/lithium/lithium-supply-demand, which seems to be based on 2016 data, "Total, an oil giant, predicts 30% penetration of car market by EVs by 2030, for 20 million sales. This would require 1-1,200,000t LCE, or 6x current global production". So they put global LCE production circa 2016 at 200,000t.
At the 48 million global EV sales by 2030, extrapolated using the forecasts in those Bank of America Merrill Lynch reports referenced in post 30533947, this would be 2,400,000 - 2,880,000t LCE or over 14x the 2016 global production levels alluded to above.
Found some conflicting historical global lithium production figures https://www.statista.com/statistics/606684/world-production-of-lithium/ shows 31500 tonnes (34723 tons) for 2015 compared to https://minerals.usgs.gov/minerals/pubs/commodity/lithium/mcs-2017-lithi.pdf which states 49,400 tons (44,815 tonnes) for 2015, on balance I'd probably tend to rely on the latter. The former also shows 35000 tonnes for 2016. At 35,000t Li converts to 186,170t LCE and 44,815t Li converts to 238,377t LCE so they do tend to indicate the 200,000t pa 2016 global LCE production figure used by Total is not unreasonable.
If the EV sales forecasts from Total are anywhere in the ballpark, that would easily absorb the 874,500t LCE to CY2020 (from totalling "Now" to "2020" inclusive in CaptainBarnacles latest spreadsheet) even without allowing for other uses lithium is needed for. IMO if the BoA-ML view of things comes to pass then there will need to be a lot higher levels of LCE production coming on stream over the next 12 or so years.
