Going Natural: The Solution To Tesla’s Graphite Problem

On February 26, 2014, Tesla Motors Inc. (NDQ:TSLA) announced details of its long-awaited “gigafactory”, an ambitious plan to build a facility to manufacture lithium-ion batteries in large-enough quantities to meet the needs of the 500,000 electric vehicles (EVs) that the company plans to produce in 2020. Tesla proposes to build this facility somewhere in the southwest United States, in reasonable proximity to its California-based vehicle assembly plant.

Tesla’s plans call for the creation of 35 GWh/year of production capacity for its third-generation Model E vehicle, implying an average 70 kWh of storage capacity per vehicle. The plan calls for an additional 15 GWh/year of production capacity, presumably to meet the needs of additional ventures with which Tesla founder and CEO Elon Musk is involved.

In addition to the significant quantities of lithium, cobalt and other metals that the batteries from this proposed facility will require, even greater quantities of graphite will be needed to produce the anodes that are used in these batteries.

It is obviously early days for the gigafactory initiative, and a number of important details have yet to emerge. There is certainly no guarantee that Tesla will actually move forward with the project, or that it might not morph into some other form. Nonetheless, it is important that the supply chain gets itself ready to participate.

Bloomberg picked up on the use of graphite in lithium-ion batteries in a March 14, 2014 article. Titled “Teslas in California Help Bring Dirty Rain to China“, Bloomberg linked the future Tesla facility to the significant pollution generated by China’s natural-graphite industry, which has “fouled air and water, damaged crops and raised health concerns“. Of particular concern is the use of acids by the Chinese industry to purify mined graphite so that it can be used in battery anodes.

Tesla’s battery supplier is Panasonic Corporation (TYO:6752), providing battery packs for the Tesla Model S vehicle that contain more than 7,100 individual 18650-model cells. It is unclear if Panasonic uses synthetic or natural graphite in these batteries, or how such materials are processed. Mr. Musk did use his Twitter account on the same day that the Bloomberg article was published, to describe the accusations as “[b]eyond ridiculous“.

Nonetheless, synthetic graphite is twice the cost of battery-grade natural-flake graphite, and is typically derived from petroleum coke, which relies on crude oil as its source. Tesla has a stated goal of reducing the unit cost of battery production by a minimum of 30% between now and the initial ramp-up of the Model E in 2017. Natural flake graphite stands to play a significant role in reducing the unit costs of battery production and in reducing the environmental footprint associated with production, if acid-based purification steps can be avoided.

The present flake-graphite market is dominated by China; aside from the issues of pollution, there is increasing evidence that the country’s flake-graphite resources are becoming depleted. Fortunately, there are a number of promising flake-graphite projects under development outside of China.

Producing battery-grade flake graphite
As with any natural resource, the quantity and grade of any given graphite deposit is important, but the type and distribution of the flake size, their purity and their amenability to processing dictate the quality of any given project.

The graphite ore is mined from the deposit and is subject to standard beneficiation processes, such as crushing, milling and flotation. The higher the initial head grade, the cheaper this process will generally be, all other things being equal. The resulting graphite concentrate, known as run-of-mine, or ROM concentrate, is typically sold directly to end-users in a number of sectors. Some junior mining companies are planning to upgrade a portion of that ROM concentrate internally for specific applications, such as the high-purity graphite used in the production of battery anodes.

Battery-grade graphite requires very high purity levels, typically >99.9% carbon-as-graphite (Cg). This material also needs to be spheroidized using careful processes that convert the flat graphite flakes into potato-like shapes, which pack much more efficiently into a given space. The high purity levels and the enhanced “tapping” density (to >0.9 kg/m³) are important for producing the high electrical conductivity that is required during anode operation.

Spheroidizing the graphite flakes also reduces their size, a process known as micronization. Standard battery-grade materials require an average diameter of approximately 10-30 μm, so in theory, feedstock materials with flake sizes greater than 30 μm (+400 mesh) could be used. However, starting purity levels tend to decrease with flake size, so flake material with an average diameter of 150 μm (+150 mesh) or greater is typically used. This is, of course, a double-edged sword, since the larger the flakes used, the more energy will be required to reduce the average size of the flakes to the desired 10-30 μm. Smaller particles are preferred, as this makes it easier for the lithium ions in the electrolyte to diffuse between graphite particles.

It should be noted that it is the tendency for purity levels to increase with flake size that is the real reason for the common ‘mantra’ that for battery-grade materials, the bigger the flake size, the better. In fact, the ideal precursor material would have small flake size if it had sufficient purity levels for the subsequent processing to be cost-effective.

One other important factor in the production of battery-grade materials is that of wastage. The standard spheroidizing and micronizing processes used in China waste up to 60-70% of the mass of total graphite flakes present during processing. Therefore, for every one tonne of spheroidal graphite produced in China, approximately three tonnes of feedstock materials might be required (though the waste materials can be used for other purposes).

The graphite may be purified before or after spheroidizing and micronizing, depending on the manufacturer. As mentioned earlier, the low-cost approach typically used in China is to leach the impurities from the graphite with acid, with the associated environmental concerns that that brings. Alternatively (and far more acceptable in Western jurisdictions), a thermal process can be applied. This typically involves the use of halogen gases to cause chemical reactions at high temperatures with the impurities, which covert the resulting compounds into gases too and eliminate them from the bulk graphite material. The higher the starting purity levels of the graphite after initial concentration at the mine site, the lower the cost will be for purification, and this can make a substantial difference when comparing concentrate feedstocks with different starting purity levels. TMR estimates that the cost difference in purifying a 95% Cg concentrate to >99.9% Cg, versus taking a 98% Cg concentrate to >99.9% Cg could be as much as $2-3,000/t of concentrate, using thermal processes.

The final step for preparing spheroidal graphite for anode production is the application of a coating to the particles to reduce their specific surface area. This is important, as reducing the specific surface area will increase the long-term capacity of the battery cell. Intercalation of the electrolyte solvent into the graphite and its reaction with it causes expansion of the graphite, with the potential for delamination and a lowering of the life expectancy.

During the first charge of the battery cell, an initial, irreversible chemical reaction occurs between the electrolyte and the graphite in the anode, resulting in the formation of a so-called surface electrolyte interphase (SEI) layer. Once formed, this layer reduces further decomposition of the electrolyte and actually protects the graphite anode from exfoliating.

With too large a specific surface area, the formation of the SEI layer can reduce the graphite’s ability to subsequently hold and to release the lithium ions in the electrolyte, thus reducing lifetime capacity for the battery. Coating the graphite prior to anode production reduces this effect and helps to maintain the maximum capacity possible for the battery. The coating can also reduce the chances of a runaway chemical reaction in the battery.

Such coatings are typically carbon- (not graphite-) based; uncoated spheroidal graphite typically sells for $3-4,000/metric tonne (t); coated spheroidal graphite sells for $9-10,000/t. The battery manufacturers typically apply these coatings, though some traders will buy uncoated materials and apply coatings before selling the finished product to the battery manufacturers.

How much battery-grade graphite will Tesla need?
Let’s return now to Tesla and its proposed gigafactory. We know that the 500,000 EVs that Tesla has planned for 2020 will require a total of 35 GWh of energy storage. We now need to determine how much graphite will be contained in those batteries.

The Department of Energy estimates that graphite constitutes approximately 16% by weight of a typical lithium-ion battery. The Panasonic spec sheet for its 18650 batteries indicates that each cell weighs 45 g, which means that the 7,104 cells in the 85 kWh battery pack for the 2013 Tesla Model S weigh approximately 320 kg. We can therefore estimate that these batteries use approximately 0.62 kg of graphite/kWh storage capacity – over 54 kg of graphite per 85 kWh vehicle. Note that the battery pack for a Tesla Model S is approximately four times the capacity of a “standard” battery EV.

This translates into approximately 21,600 t of graphite required for the 500,000 batteries (each with 70 kWh capacity) needed in 2020. However, we need to account for the relatively low (30%) yield of battery-grade graphite, using current processing methods. This means that some 72,050 t of graphite feedstock would actually be required for these batteries at those yields.

Using today’s prices for synthetic (~$20,000/t) and coated spheroidal natural graphite (~$10,000/t), all other things being equal, a switch from all-synthetic to all-natural-graphite anodes for those 500,000 EVs/year would save $216M in material costs, which translates to over $6/kWh, or over $430 per vehicle. Not a bad start.

On top of the batteries for its EVs, Tesla will need a further 9,250 t of graphite for the additional 15 GWh/year of non-EV capacity at the gigafactory, which in turn, would require 30,900 t of graphite feedstocks for the production of battery-grade materials, at current yield levels.

This is a total of just under 30,900 t of graphite in the batteries, requiring 102,900 t of feedstocks using current processing methods and yields. This is over 125% of the global natural flake graphite market, currently at 80-85,000 t/year!

Who can supply this battery-grade graphite?
Clearly, there is a potential significant imbalance between current levels of supply and the projected future demand for graphite, if the Tesla gigafactory comes on-stream.

TMR tracks graphite projects under development via the TMR Advanced Graphite Projects Index. The minimum requirement for a project’s inclusion on the Index is for it to have an NI 43-101- or JORC Code-compliant mineral resource estimate. At present, there are 23 such mineral resources on the Index associated with 20 graphite projects being developed by 16 different companies in 8 countries. Those resources are:

Projects on the TMR Advanced Graphite Projects Index (March 2014)

Any number of these projects potentially has what it takes to become successful graphite mines, especially given the pressure that demand from lithium-ion batteries and other applications might put on the overall supply chain. However, to be in a position to be able to produce battery-grade graphite by 2017, when Tesla says that it will commence production ramp up of the Model E, only projects that are far enough along are likely to have the opportunity to capitalize on the demand from the gigafactory. One can argue as to how to define “far enough along,” but my suggested requirements would include:

• The project should have as a minimum, Demonstrated mineral resources (i.e. Measured + Indicated);
• The project should have a completed Feasibility Study *FS), or have one underway;
• A purification process for getting to battery-grade (>99.9% Cg) should have been defined and successfully tested (preferably without using the wet acid method); and
• A spheroidization and micronization process should have been defined and tested.

Additional considerations relate specifically to potential costs of production, and include:

• Initial grade of in-situ graphite (relates to beneficiation costs);
• The resulting purity levels of the resulting run-of-mine (ROM) concentrates after beneficiation (relates to subsequent purification costs);
• The proportion of smaller flake materials with higher purity levels after beneficiation (related to subsequent spheroidization and micronization costs and yield levels);
• Whether or not a coating process has been developed and tested for the spheroidal graphite; and
• Proximity of the project to the southwest US, proposed home of the Tesla gigafactory.

I acknowledge that there can be no ‘definitive’ list of criteria for assessing projects for this exercise, but nevertheless, the above are what I’ve chosen to use. I re-iterate that I am looking here only at the potential ability of graphite projects to service the needs of the Tesla gigafactory within the announced time frame for its development. There are plenty of other future opportunities for companies and projects that might not quite be ready for the gigafactory.

Applying the initial set of criteria to the projects on the TMR Index results in the following three companies (in alphabetical order) and their projects:

• Focus Graphite Inc. with the Lac Knife project in Quebec, Canada;
• Northern Graphite Corp. with the Bissett Creek project in Ontario, Canada; and
• Syrah Resources Ltd. with the Balama project in Mozambique.

It should be noted that additional companies may be developing high-purity, spheroidized and micronized battery-grade graphite, but to my knowledge, these are the only three that have discussed such developments and their achievements in the public domain, and which meet the other primary criteria.

Focus Graphite Inc.
Focus announced the completion of a Preliminary Economic Assessment (PEA) in October 2012 on its Lac Knife project in Quebec, Canada. In November 2013, the company updated the economics for the PEA, which put the capex requirement for going into production at $126M (including a $24M contingency). It also announced commencement of a definitive FS for Lac Knife, which is slated for completion by late spring or early summer this year.

Lac Knife has total Demonstrated mineral resources of 9.6 Mt @ an average 14.8% Cg. Focus plans to produce 44,300 t/year of graphite from the future mine, with a mine life of 20 years. Metallurgical work to date indicates that the ROM concentrate will have a purity level of >96.6% Cg and will cost a total of $458/t to produce.

The flake graphite in the Lac Knife ROM concentrate is distributed as follows:

The >98% Cg purity levels of the Lac Knife flake above 75 μm (constituting almost 80% of the content) is particularly high. The company recently indicated that this is a result of most of the impurities being found at the surface of the flakes, instead of being “ingrained” in the layers.

In November 2013, Focus announced that it was working on the production of spheroidal graphite from Lac Knife concentrates and the development of purification processes for producing battery-grade graphite. During this month’s PDAC Convention in Toronto, Focus showed samples of 99.95% Cg battery-grade, spheroidal graphite. Company management has subsequently indicated that coatings for this battery-grade material are currently being tested, the results of which should be announced in the near future.

In an industry first, Focus announced a significant off-take agreement in December 2013 with a Chinese industrial conglomerate for up to 40,000 t/year of its concentrates. A clarification earlier this month indicated that this agreement calls for a minimum purchase of 20,000 t/year by this Chinese group.

Northern Graphite Corp.
Northern announced the completion of a definitive FS in July 2012 on its Bissett Creek project in Ontario, Canada. In August 2013, the company received final approval for the project and was granted a mining lease, allowing it to begin construction subject to financing. In September 2013, the company updated the economics for the FS, which put the capex requirement for producing 20,800 t/year, with a mine life of 28 years, at $101.6M (including a $9.3M contingency). Operating costs were estimated at $795/t ROM concentrate.

In October 2013, Northern announced the completion of an “Expansion Case” PEA for the Bissett Creek project, which would see an increase in the production rate to 33,183 t, an initial capex of 146.8M and reduced operating costs of $695/t.

Bissett Creek has Probable mineral reserves of 28.3 Mt @ 2.1% Cg. Total Demonstrated mineral resources are an estimated 69.8 Mt @ an average 1.7% Cg. Metallurgical work to date indicates that the ROM concentrate will have a purity level of >96% Cg.

The flake graphite in the Bissett Creek ROM concentrate is distributed as follows:

In October 2012, Northern announced that it had successfully produced spheroidal graphite from Bissett Creek concentrates, with up to 70% yields when starting with so-called large (180-300 μm or -50 / +80 mesh) flake. In September 2013, the company announced the development of a proprietary method for the purification of concentrates and spheroidized graphite to 99.95% Cg. The company claims that the cost of this purification process will be less than $1,000/t. Company management indicates that this is a “low-temperature thermal process” that uses no acids, in which a mixture of gases, tailored to the impurities and mineralogy of the Bissett Creek deposit, is used.

In November 2013, Northern announced that it had partnered with Coulometrics to develop coatings for their spheroidal graphite, and earlier this month, the company announced the completion and successful testing of this work.

Syrah Resources Ltd.
Syrah announced the completion of a Scoping Study (SS) for its Balama West deposit in June 2013, one of a number of endeavors associated with its Balama project. However, because the study was based on an Inferred mineral resource only, the company was required to clarify that the SS did not demonstrate economic viability. I have been unable to find many more details or numbers in an announcement on the SS itself. Balama West was subsequently upgraded, with reported total Demonstrated mineral resources of 13.6Mt @ 19.8% Cg.

A statement in December 2013 indicated that the company was in the process of undertaking an FS to be completed by Q1 2014.

In January 2014, Syrah announced that it had purified ROM concentrate to >99.9% Cg using a “chemical wash” containing acids. Earlier this month, the company announced that it had produced spheroidal graphite with an average diameter of 4.7 μm, from a 100 μm feedstock. It is unclear as to why the graphite was micronized to below the standard 10-30 μm range typically used for battery anodes.

The flake graphite in Balama East ROM concentrate is distributed as follows [per the same announcement made earlier this month]:

It should be noted that the published flake size ranges are for Balama East (which does not have Demonstrated mineral resources), as opposed to the Balama West deposit.

In March 2014, Syrah announced the completion of a Memorandum of Understanding (MOU) with a subsidiary of Chinalco Group of China for an off-take of 80-100,000 t of graphite over an unspecified time period. The MOU required the two parties to negotiate a binding off-take agreement within three months of execution.

Discussion
From their public announcements, it is clear that these three companies are well on their way to achieving the technical milestones required for the production of battery-grade graphite.

In terms of other parameters, however, it may be a little early to determine if Syrah Resources will be able to provide such materials in time for the launch of the proposed Tesla gigafactory. The lack of a detailed, published SS summary or other quantitative analysis makes it difficult to evaluate the applicability of the Balama project to the gigafactory at this time. Completing a binding off-take agreement with the Chinalco subsidiary could go some way to assuaging such uncertainty; however, the larger issue probably relates to the location of the deposit in Mozambique – a long way from southwest United States. This distance puts Balama at a disadvantage in terms of transportation costs when compared to the other two projects, and potential access to infrastructure.

In terms of technical progress, the results announced by Northern Graphite to date indicate that it has a product that is ready to go for the battery market, with similar definitive results expected from Focus Graphite in the very near future. Based on the Bissett Creek Expansion Case PEA and the Lac Knife PEA (and subsequent updates), Focus appears to have significantly lower operating costs to produce its ROM concentrate, and the resulting concentrate has higher Cg purity levels across the various flake-size ranges.

The projects of both of these companies are located in Canada, which presents relatively inexpensive transportation costs for getting materials to the southwest United States.

It is always hard to compare ‘apples to apples’ when it comes to graphite project flake size, as each company has its own mesh size ranges that it uses to report. However, it appears that the Lac Knife ROM concentrate has a significantly higher proportion of flake, at higher purity levels after beneficiation than that for Bissett Creek. Northern states that the cost to get to high-purity battery grade with its proprietary process is less than $1,000/t; how much less will determine the overall costs required to get from the lower purity levels in the ROM concentrate to 99.9% Cg. Starting at levels above 98% Cg, as Lac Knife does, is a distinct advantage.

The completion of a definitive FS for Northern’s project means that the costs of production of ROM concentrate that result from the study have a higher degree of certainty at this point than those detailed in the Lac Knife PEA (and subsequent updates) for Focus Graphite. Relying on the Expansion Case PEA lowers the production-cost estimates for Bissett Creek, but decreases the accuracy of those numbers compared to the original FS.

It remains to be seen if there will be any increase in anticipated product costs in the forthcoming Lac Knife FS. The costs would have to increase by more than 50%, however, to reach those estimated in the Bissett Creek Expansion Case PEA. The Lac Knife cost advantage is likely a result of the significantly higher grade to be found at Lac Knife (14.8% Cg for Demonstrated resources vs. 1.7% Cg at Bissett Creek), and the proposed larger annual production rate, which naturally gives economies of scale. One might question why the estimated cost differential is not actually higher, given the distinct baseline differences, but it likely indicates that the cost of physical comminution (crushing and grinding) of the graphite ore is a relatively small fraction of the overall mining and processing costs.

Finally, there can be no mine without the financing to construct and to operate the mine. While Northern is no doubt negotiating with third-parties on future off-take agreements and other arrangements, the fact that Focus was the first company in the sector to complete such an agreement, and for such a substantial proportion of its planned output, is significant.

Based on the above criteria, I believe that both Northern and Focus have the potential to take advantage of the Tesla gigafactory, if it comes to fruition. Given the volumes of material required, it is entirely possible that both could service the supply chain requirements. I believe, however, that Focus may well stand to gain greater benefit from the opportunity that the gigafactory presents. Its lower ROM concentrate operating costs, likely lower battery-grade purification costs and the fact that it has already secured a significant off-take agreement (making it that much easier to finance the eventual construction of the mine) are key factors.

I wish both companies well, and I make one last observation. Even if the Tesla gigafactory does not come to pass, there will undoubtedly be opportunities for these and other graphite companies to service the needs of the wider EV market in the years to come. Doing so does not need to come at the expense of the environment either, as appears to be the case in the Chinese graphite sector.

Vehicles ‘fueled’ by electricity, especially electricity generated via renewable means, need to be built using as sustainable and environmentally-friendly a supply chain as possible. In the case of batteries for EVs and the graphite required to make them, natural-flake sources are clearly the way to go.

Disclosure: at the time of writing, Gareth Hatch is neither a shareholder of, nor a consultant to any of the companies mentioned in this article.