Cobalt – A Coffin Nail for Cheap Lithium-ion Batteries

In early March, I learned that only 6% of the world’s cobalt is produced as a primary product while 94% is a by-product of nickel and copper mining. I found the ratios alarming because:

  • Cobalt is an essential raw material for all high-energy lithium-ion batteries;
  • While conservative analysts believe the battery industry’s cobalt requirements will double over the next 10 years, rapid and sustained growth in electric vehicles (EVs), stationary energy storage and other applications could drive the multiple much higher;
  • By-product availability is always dependent on sales of the primary metal and miners cannot respond to normal market signals for by-product metals;
  • With nickel and copper trading at multi-year lows, international mining conglomerates are suspending operations at marginal mines around the world; and
  • Due to these mine closures, cobalt production is expected to fall in 2016 and remain low until nickel and copper prices recover and mines return to production.

As I noted last week, the battery industry’s apparent failure to understand supply and demand dynamics in the cobalt market is a due diligence debacle that may well be the biggest “oops” in the history of supply chain management.

While my earlier articles suggested that most competitive users of cobalt are in a better position than the battery industry when it comes to weathering price volatility and defending their supply chains, I didn’t offer granular historical data to support my conclusion.

I am providing that data in this article.

Metal Production Data

One of the more confusing aspects of metal production data is that industry-standard reporting practices use three different metrics.

  1. Mine production data, a metric used by the U.S. Geological Survey (USGS), represents the total metal contained in ore concentrates shipped from mines during a year.
  2. Refined metal production data, a metric used by the Cobalt Development Institute (CDI), represents the total reported production of refined metal during a year.
  3. Apparent demand, a metric used by the World Bureau of Metal Statistics (WBMS), represents the total amount of refined metal shipped to customers during a year as calculated from import/export data.

In general, mine production is the largest number because it includes all metal values in shipped ore concentrates. Refined metal production data is usually a smaller number because smelting and refining invariably give rise to processing losses, refined metals are not 100% pure and the time lag between when a concentrate is shipped from a mine and when the contained metals are refined can be lengthy. Apparent demand is usually the smallest of the three values because it focuses on actual deliveries of refined metal to customers. I want to thank David Weight, the CDI’s president, for his willingness to walk me through the differences between these metrics and his patient responses to my other questions.

The following table summarizes the three values for cobalt for 2006 through 2014.


Industrial End Use Data

In addition to statistics on refined cobalt production, the CDI also publishes annual estimates of the percentage of apparent demand absorbed by participants in nine discrete industrial sectors. The sectors are described in detail in the cobalt supply and demand section of the CDI’s annual Cobalt Facts. The end-user demand estimates are, by nature, approximations based on available CDI and WBMS statistics. They’re certainly close enough to provide a reasonable overview of historical demand conditions in the cobalt market.

The following graph takes annual apparent demand statistics from the WBMS and uses industry absorption percentages published by the CDI to calculate total cobalt shipments to each sector tracked by the CDI. The red line in the background comes from and shows how market prices for cobalt fluctuated between $10 and $52 per pound during the nine-year period from January 1, 2006 through December 31, 2014.


When I consider the tonnage of cobalt shipped to the individual non-battery industrial sectors each year, I’m surprised by the consistency, modest year on year growth rates and insensitivity to market prices. It certainly appears that each of the non-battery sectors has a fixed and relatively predictable annual requirement that they must satisfy regardless of price. An economist friend observed that demand from the non-battery industrial sectors looked to be extremely inelastic.

When I consider the tonnage of cobalt shipped to battery manufacturers during each of the nine years ended 2014, I’m amazed that 2014 demand was 270% of 2006 demand. Since the battery industry is extremely sensitive to raw material prices, I’ll simply observe that a return to $52 a pound for cobalt would add $25/kWh to the bill of materials for NCA cells; $41/kWh to the bill of materials for NMC cells; and $164/kWh to the bill of materials for LCO cells.


Cobalt’s status as a by-product of nickel and copper mining precludes the possibility of adequate cobalt supplies for the battery industry. Moreover, competition from well-established industrial users of cobalt precludes the possibility that recent low prices will continue in the future despite declining global production and increasing global demand.

Something has to give.

Based on everything I’ve learned about supply and demand dynamics in the cobalt market over the last month, I’m convinced the Cobalt Cliff will be a catastrophic “perfect storm” for lithium-ion battery manufacturers, EV manufacturers and developers of stationary energy storage products that can’t be cost-effective without plentiful supplies of dirt cheap batteries.

Have Your Say!