David Fickling is a Bloomberg Gadfly columnist covering commodities, as well as industrial and consumer companies. He has been a reporter for Bloomberg News, Dow Jones, the Wall Street Journal, the Financial Times and the Guardian.

(Corrected )

Does the world have enough lithium? It depends who you ask.

A 2008 study by French researcher William Tahil found there were just 3.9 million metric tons of recoverable deposits globally in mineral ores and Andean salt lakes. That's little enough that the world would risk running out as demand for lithium-ion car batteries and utility-scale storage ramps up over the coming decades.

Take Your Pick
Estimates of global lithium reserves vary by a factor of 10
Source: Mohr, Mudd & Giurco, "Lithium Resources and Production: a critical global assessment" (2010)

A survey the following year by consultants Gerry Clarke and Peter Harben, though, concluded there was about 10 times that amount. Depending on the other parameters applied, those numbers suggest deposits could provide lithium for anything from a further 100 million cars -- about 10 percent of the global auto fleet -- to 10 billion or more.

That vast range of estimates is inevitable given compounding uncertainties around the quantities available and the amounts needed. Still, a glimpse of how the sausage is made may help to demonstrate why, as we argued in Wednesday's column, fears of peak supply for battery materials should be taken with a pinch of altiplano salt.

Calculating mineral reserves is a little like working out the contents of a lucky dip by sticking knitting needles into a box. Geologists must drill through rock to produce kilometers of core samples at a cost of several hundred dollars a meter, and then analyze what they've dug up to work out the rough shape and concentration of an irregular ore deposit deep below ground -- a mineral resource.

After that, engineers assess the cost of building and running a mine and processing plant, while economists estimate future prices. Put that financial and economic data together with the geological information and you have a mineral reserve, the gold standard for working out supply availability.

As should be evident, though, that estimate contains multiplying layers of variables. A movement in just one can, for example, make 3.5 billion barrels of oil disappear. That's one of the reasons people invest in mineral exploration companies: There's always the chance that a few extra bits of data could produce a dramatic change, as when a 2007 update increased the estimated lithium content of Argentina's Salar del Rincon brine lake from 253,000 tons to 1.4 million tons. Geologists' estimates of Rincon's lithium concentration range from 0.033 percent to 0.05 percent, a difference that on its own would be enough to increase the resource base by 50 percent or reduce it by a third.

Garbage In, Garbage Out
Whether the world runs out of lithium depends a lot on your assumptions
Source: Bloomberg New Energy Finance; Citigroup; Deutsche Bank; the companies
Note: This chart models the differences that small changes to assumptions can make and shouldn't be taken as a forecast. High consumption is based on an average 85kWh battery with 2kg/kWh of lithium; mid is 60kWh battery with 0.7kg/kWh; low is 40kWh battery with 0.7kg/kWh. Assumptions for size of car fleet based on BNEF forecast.

Then come demand issues. The amount of lithium used per kilowatt-hour of battery power is changing rapidly as manufacturers improve efficiency. One 2015 study by Citigroup Inc. estimated about two kilograms of lithium carbonate equivalent are needed for each kWh; more recent appraisals by Deutsche Bank AG and Advantage Lithium reckon it's about a third as much.

On top of that, there's the question of how powerful batteries will be. A hybrid Toyota Motor Corp. Prius has a 4.4kWh battery; a Tesla Inc. Model S can have one as large as 85kWh. Most analysts base their estimates on the 30kWh to 40kWh range of Nissan Motor Co.'s all-electric Leaf, but in truth it's anyone's guess which size will dominate.

There's another issue. Most analysts expect global lithium demand to double, or even triple, by 2030 -- but stronger usage by the battery sector may cause other industries to seek alternative materials. For instance, only in the last few years have rechargeable batteries started to consume more lithium than glassmaking, where sodium and potassium compounds may be effective substitutes. As recently as 2007, lubricating grease was almost as important an end-use sector as batteries. As prices rise, those industries may find alternative chemicals work as well.

Cell Division
Rechargeable batteries have only recently become the biggest end-use sector for lithium
Source: U.S. Geological Survey
Note: Based on estimates of global end-use sector share.

Further off, alternative battery chemistries -- based, for instance, on sodium and silicon -- may cause lithium-ion cells to go the way of the nickel-iron cells that Thomas Edison hoped would power a previous generation of electric cars.

None of this guarantees that lithium won't run short or that prices won't rise for brief periods in the near term. If Tahil is right and demand grows fast enough, the market could get awfully tight. Those hoping for a dearth, though, need to reckon with the possibility that we could be hurtling towards a glut, instead.

Second of a three-part series on battery materials. Part one was on platinum and graphite, and part three, on cobalt, appears tomorrow.

This column does not necessarily reflect the opinion of Bloomberg LP and its owners.

(The seventh paragraph of this story was corrected to show that the upper range of estimates for Rincon's lithium concentration is 0.o5. percent, not 0.5 percent.)

  1. Technically, lithium carbonate equivalent, the battery industry standard. Miners, confusingly, prefer to enumerate their reserves in terms of lithium metal instead. There's about 188 grams of lithium metal for every kilogram of lithium carbonate equivalent.

To contact the author of this story:
David Fickling in Sydney at

To contact the editor responsible for this story:
Matthew Brooker at