Platinum and hydrogen for fuel cell vehicles

2  Platinum

Platinum is a precious metal belonging to the Platinum Group Metals (PGMs) widely used in the automotive, jewellery, chemical and electrical industries. Its properties allow it to be used as a catalyst in conventional petrol and diesel catalysts, and also in fuel cell stacks. Commonly, platinum is alloyed with other PGMs such as Rhodium or Palladium to form catalysts. However, platinum is used on its own in hydrogen fuel cell stacks and in diesel engined autocatalysts. The current demand for PGMs is greater than the supply, but historically supply has risen to match increasing demand. This supply is sourced from only a few countries worldwide, and as such, the reserves are geographically limited. The anticipated introduction of fuel cell vehicles has raised many concerns about the future availability of platinum and other PGMs. The mass penetration of fuel cell vehicles as part of the world fleet could have a serious impact on the demand for PGMs, and raises the question of the potential availability of alternative metal catalysts for both fuel cells stacks and autocatalysts. A further constraint is the cost, which is currently about £450 per ounce. Platinum is a traded commodity and, as such, its prices are subject to major annual and monthly price fluctuations.

Platinum and other PGMs are usually quantified in ounces (oz) or troy ounces (troy oz), where 1 oz = 28.3 g and 1 troy oz = 31.1 g. We have tried to use ounces throughout this report for consistency, but note that the literature is not always clear on which units are being used.

2.1 Platinum Loadings

How much platinum does a car-sized fuel cell stack currently use?

How is this likely to change by 2020 assuming mass production of stacks?

How did platinum loadings change over time for 3-way automotive catalysts?

How did this differ from predicted reductions?

Fuel cell platinum loadings have fallen dramatically in recent years, with a reduction of a factor of ten between 1994 and 1999 (Cahn, 2002). According to General Motors purchasing director David Andres, fuel cell stacks for cars currently use about 2 oz of PGM per unit. Pure platinum catalysts are used for hydrogen-fuelled fuel cells, while alloys of platinum with ruthenium are typically used for reformed hydrocarbon fuels to improve the tolerance of the catalyst to carbon monoxide.

Fuel cells. The fuel cell research team at Johnson Matthey believes that loadings can be reduced to about 1 oz by 2010 through better utilisation of platinum and thinner deposition layers (Potter 2002). Other experts at Johnson Matthey and Advanced Power Sources estimate that when fuel cells are commercially produced each engine will require between 0.2 and 0.3 oz of platinum (Evans 2002; Uravan Minerals 2002; Adcock 2000). The lowest laboratory data on platinum loading for efficient performance of fuel cells is currently about 0.8 oz (Borgwardt 2001).

Catalysts. Tighter vehicle emissions controls have meant that PGM loadings for petrol and diesel catalysts are increasing. There is evidence that future emissions legislation in the USA, Europe and Japan will drive even higher PGM loadings (Tom & Das, 2001; Resource Opportunities, 2001), but it is difficult to obtain quantitative projections. Based on current estimates of supply, cost and the amount of PGM needed for the world fleet of cars subject to low emissions regulations, General Motors engineers have been instructed to estimate future loadings of no more than 0.048oz platinum and 0.096 oz palladium (Automotive News 2002).

Note that the above figures for PGM loadings assume a conventional European style car with an engine size of 50-80 kW. Sports Utility Vehicles (SUVs) and light trucks favoured in North America may have much higher platinum requirements.

2.2  Platinum Demand

What is the present world platinum consumption?

What is platinum currently used for?

What is the projected demand for platinum for current and prospective applications, including stationary fuel cells?

Historical and Current Demand

Table 2.1 and Figure 2.1 show trends in the worldwide demand for platinum. The two main applications are autocatalysts and jewellery. Other applications, in descending order of current demand, are the electrical, chemical, glass and petroleum refining sectors.

Johnson Matthey 2002

Figure 2.1: Worldwide demand for platinum by application, 1976-2001 (from CPM Group, 2001)

Each application and its recent demand trend are described below.

• Autocatalysts. The main consumer of world PGM supply is the automobile industry. 41% of platinum demand in 2001 was accounted for by autocatalyst use. (Hilliard 2001). PGMs are used in autocatalysts to facilitate the removal of three of the main combustion by-products; CO, hydrocarbons and NOx. Data from Johnson Matthey shows that in 2001 demand for platinum in autocatalysts was 2.52 million oz. In Europe, consumption of platinum for autocatalyst use increased. The use of platinum increased by 375,000 oz in 2001 due to strong growth in production and sales for diesel cars. Diesel autocatalysts only use platinum rather than the mixtures of platinum and palladium commonly used in petrol autocatalysts. In 2001, diesel catalysts accounted for over 70% of auto demand for platinum in Europe (Johnson Matthey plc 2002).
• Jewellery. Platinum jewellery is increasing in popularity across the world. Typically the demand has been strong in Japan, the US and Western Europe. China is now showing major demand. Demand for platinum jewellery dropped by 10% in 2001, associated with economic downturn in Japan and the US. However, Chinese demand increased. Platinum jewellery is rarely manufactured using pure platinum. More often it is strengthened by having it alloyed. Platinum purity of a piece of jewellery is measured in parts per thousand. In Western markets, the jewellery market is dominated by the bridal products which are generally less sensitive to price increases. Total demand in the jewellery sector was 2.55 million oz in 2001.
• Chemical. Platinum is used in the manufacture of silicone. As most of the platinum is lost during the production process. Silicone is important in the manufacturing of adhesives, synthetic rubber, sealants and a range of personal car products. Platinum demand is closely related to silicone output. Platinum use in other sectors of the chemical industry is small compared to silicone. A small amount is used in the fertilisers. The total demand last year was 290,000 oz.
• Electrical. Computer hard disks are the largest single electrical application for platinum, and demand almost doubled between 1998-2000. Over 90% of computer hard disks use a platinum alloy layer (Johnson Matthey plc 2002). A small amount of platinum is used in thermocouples. Consumption in the electrical sector in 2001 was 385,000 oz, representing a fall of 15%, attributed to a sharp downturn in the global electronics industry.
• Glass. Glass manufacturing uses platinum. The demand worldwide was driven by growth in China for new televisions and investment in Japan and Korea of LCD glass. The total demand in this sector was 285,000oz in 2001.
• Petroleum refining. Petrol refineries use platinum based catalysts in several stages of the refining process, particularly the reforming of naphtha into downstream products (Johnson Matthey plc 2002). The consumption in Europe, China and Japan was stable last year, though is forecast to rise in Latin America and South East Asia. There was an overall increase in demand in this sector to 125,000 oz in 2001.
• Other applications. Platinum alloys are used in the dental industry as replacement fillings. Oxygen sensors for cars, sparkplugs, catheters and pacemakers all contain PMG alloys. The demand for other applications rose by 16% in 2001 to 435,000 oz.

Future demand

Projected demand for platinum depends on a number of different parameters; estimated platinum reserves, platinum demand factors, platinum supply factors, penetration rate of fuel cell vehicles and platinum recycling rates. Historically supply has met demand as new resources are exploited on the basis of expectations of demand growth.

Platinum demand projections expect there to be a significant increase in the use of platinum for current automotive technology. Development of fuel cell vehicles, and stringent vehicle emissions standards has led to the development of more advanced autocatalysts for both petrol and diesel cars. J.M Christian (1999) shows that consumption of platinum for autocatalysts is projected to rise by 5% each year. The current market penetration of diesel engined cars has important implications for the future use of PGM. In some parts of Europe, diesel cars have penetrated over 50% of the market (Johnson Matthey plc 2002). Fuel cell vehicles are expected to start to penetrate the market within the next 5-10 years, hence demand for platinum will increase.

Fuel cells are also being considered for stationary applications such as power generation, CHP and portable power (laptop computers etc.). Large-scale power generation applications are likely to use higher temperature fuel cell types such as Molten Carbonate Fuel Cells or Solid Oxide Fuel Cells, neither of which use platinum in the stack. However, platinum-containing fuel cells are being considered for small-scale power generation, CHP and portable power. If these lower temperature fuel cell systems were to provide 10% of new power generation equipment sales by 2010 then this would equate to about 14 GW per year (DTI 1997), or about 0.1 million oz/year of platinum assuming platinum loadings are about twice those required for transport applications. Loadings for stationary applications are likely to be higher as the stacks operate for many more thousand hours and catalysts are degraded by hydrocarbon fuels. Future markets are very uncertain but the mass deployment of fuel cell vehicles is likely to boost the technology and bring down prices such that fuel cells are also attractive in stationary applications. Typically the cost targets for fuel cell engines are about £400/kW for stationary applications and below £50/kW for transport.

Projected demand for platinum and its relationship with the rate of penetration of fuel cell vehicles is explored further in Section 2.7 of this report.

2.3  Platinum Supply

Where does platinum come from?

Are there accessible reserves in other countries?

How easily could production be expanded?

How would a major expansion of production affect the price of platinum?

South Africa and Russia dominate world supply of PGMs, as shown in Table 2.2. South Africa holds about a 70% share of the world supply of platinum and Russia about 22%. The remainder comes from Canada, the US and Zimbabwe. In 2001, South African supplies of platinum rose from 3.8 million oz to 4.1 million oz (8% increase) as producers added to output. Russian supply increased from 1.1 million to 1.3 million oz (18% increase) and North American supplies increased by from 285,000 oz to 350,000 oz (23% increase). The total world supply for 2001 was 5.86 million oz up from 5.29 million oz in 2000, representing an overall increase of 8%.

Table 2.2: Platinum supply by region in 2001

Region	'000 oz
South Africa	4,100
Russia	1,300
North America	350
Zimbabwe and others	110
Total	5,860

Historical trends in supply and demand for platinum are shown in Figure 2.2. The two have generally tracked each other closely over time, but demand has outstripped supply in recent years.

  Figure 2.2: Supply and demand of platinum, 1976-2001 (from CPM Group, 2001)

Current supply pressures are likely to ease as South African mines increase output, and Russian exports increase. Long-term supply is more uncertain. The market for platinum group metals (PGM) is very volatile resulting in cyclical price patterns for each of the main PGM metals. The US Geological Survey estimates that world reserves of PGM stand at 3.21 billion oz (Tonn and Das 2001). The majority of platinum reserves are found in the Bushveld Complex, South Africa. Cawthorn (1999) has investigated these reserves and concluded that that 939 million oz is likely to be present within 2km of the surface and mining below this level could be possible in future. He concludes that production in South Africa could be expanded at a rate of 5% per year for at least another 50 years. A major reserve estimated at 139 million oz can be found in the Great Dyke of Zimbabwe, but political problems in the country restrict mining activities. Estimated reserves in the Noril'sk deposit in Russia are 50 million oz. In addition there is by-product output from ores in the US, Canada and Peru (Blair 2000). There are also some deposits in other countries such as Australia and Greenland, but these are not economic at current platinum prices (Evans 2002).

The price of platinum has fluctuated over the last decade, as shown in Figure 2.3, but actually fell during 2001. The price appears to depend on market expectations of future supply and demand. There is also a relationship with the price of other PGMs. For example, the peak in 2000 was largely due to the substitution of platinum for palladium in some catalysts due to soaring palladium prices (Johnson Matthey plc, 2002). Any supply squeeze is expected send prices of platinum higher, and any pre-emptive buying for automotive purposes in the event of an economic turnaround could force prices still higher (Mineweb 2002).

 Figure 2.3: Price variation of platinum between January 1990 and October 2001 (from CPM Group, 2001)

It is difficult to comment on the likely platinum cost implications of increased demand for autocatalysts or fuel cells. This will be dependent on a number of factors such as industry structure, market pricing mechanisms, industry's ability to temporarily absorb costs in order to promote future markets, and the future supply and demand of other PGMs, particularly platinum. Many of these issues are outside the scope of this report but it is safe to say that supply constraints may cause platinum prices to rise in future.

Security of supply concerns may be alleviated somewhat by the fact that platinum is increasingly sourced "above ground", i.e. an increasing proportion of platinum supply comes from recycled autocatalysts rather than mining (Potter, 2002). This recycled material is owned by automotive companies who are required to take responsibility for end-of-life vehicles in many countries. This should lend stability to the market and reduce the potential impact of political or economic changes in South Africa or Russia.

2.4  Environmental Impacts of Platinum

What are the environmental impacts of platinum extraction and production?

How many tonnes of ore must be mined to extract a tonne of platinum?

What is the energy and CO2 burden per tonne of platinum extracted?

According to Friends of the Earth 12.7 tonnes of ore is extracted for one ounce of platinum (FoE 1996), and jewellery companies give a slightly lower figure of 9.1 tonnes ore per ounce of platinum.

Water and air pollution are both serious problems associated with the mining and processing of platinum. Flotation plants recycle water through several processes before releasing back into the groundwater system. Pollutants then become concentrated in a very small volume of water. Smelting of ores releases sulphur dioxide into the atmosphere. Ammonia, chlorine and hydrogen chloride gases are all released as process emissions. Most significantly, effluent is left at the end of the process. This is the base metal liquor containing iron and zinc, which is removed and the residue metals are precipitated and then landfilled. The presence of the base metal precipitate in landfill can cause groundwater to be contaminated and add to landfill leachate problems. In addition, this process generates sulphates which are converted to sodium sulphate. Some is sold to glass and detergent manufacturers though due to a fall in price, most of it is simply dumped (International Centre for Science and High Technology 2002).

The environmental impact of mining and processing operations is dependent on the grade of ore and the processes used. South Africa is more environmentally aware and uses lower impact processes and remediation. Russia has traditionally been less environmentally aware but the situation is improving (Evans 2002).

A German study (Okoinstitute 1997) on the environmental footprint of autocatalysts found that the mining and processing of each gram of platinum required 23.76 KWh of electricity and 10.45 MJ of natural gas. Based on average emission factors for South Africa (Eskom 1999), this equates to 6.4 kg carbon per gram, or 180 kg C per ounce of platinum. A current fuel cell car contains 2 oz platinum, so the processing of platinum for the fuel cell stack would give carbon emissions of 360 kg C per car. If the target platinum loading of 0.2 oz is achieved, this would fall to 36 kg C per car. This compares with lifetime carbon emissions of about 6,750 kg C per car for a diesel car with an efficiency of 1.392 MJ/km, an average distance travelled of 20,000 km/yr and a lifetime of 12 years.

2.5  Platinum Recycling

How cost-effective is the recycling of platinum from fuel cell stacks likely to be?

What are the energy and CO2 implications of platinum recycling?

The platinum in fuel cell stacks is not currently being recycled in commercial operations (McKinley 2000) though both the DOE and USGS agree that fuel cells will be recycled cost-effectively eventually. Platinum is currently recovered from autocatalysts and the processing is likely to be easier for fuel cell stacks (Potter, 2002).

Two options can be considered for the fuel cell stack:

• Re-use of the stack
• Recovery and recycling of platinum

The re-use of the fuel cell stack is not possible: improvements in flow field design over the product life mean that bipolar plates ten years old will be 'out of date', even in the absence of degradation.

The USGS suggest that fuel cells would be recycled in a way similar to catalytic converters where the platinum is separated from the stainless steel so both metals can be recycled. USGS also predicted that fuel cells would be recycled on a toll basis, where the consumer retains ownership of the valuable metals. The materials balance in existing toll recycling is better than 98% (only 2% of the metal is lost in the closed loop).

Platinum recycling from fuel cell stacks will comprise three process steps: incineration to assay a sample of platinum, smelting and solvent extraction. It may also be necessary to remove the bipolar plate and electrolyte membranes from the fuel cell stack to increase the concentration of platinum and avoid hydrogen fluoride emissions from incineration (Handley 2000).

In a life cycle analysis of the solid polymer fuel cell for passenger cars: the primary energy demand is reduced by a factor of 20 and the sulphur dioxide _{emitted is decreased by a factor of 100 and when the platinum is recycled as opposed to mining from the ore (Pehnt 2000). This is illustrated in Figure 2.4. The resultant saving in energy would make the recycling of platinum cost effective at current platinum and energy prices.}

A 70 kWe PEMFC stack currently contains approximately 2 oz of platinum but the platinum loading could drop by a factor of ten over the next twenty years (see Section 2.1). Handley (2000) believes that platinum recycling from fuel cell vehicles would still be economically viable even at 0.2 oz per stack.

Figure 2.4: Relative energy and sulphur dioxide impacts of primary and recycled platinum (Pehnt 2000).

2.6  Alternatives to Platinum

Are there alternatives to platinum and how to they compare in terms of demand, supply, cost, environmental impacts and recyclability?

In the longer term, do non-rare metal substitutes offer any real prospects?

The main alternative to pure platinum catalysts, are those using alternative PGMs. The main competitor to platinum is palladium which is sold at approximately the same price per oz as platinum, and which is closely linked in market terms. Prices of both have remained high since about 1997 when there were delays to Russian exports, which is the primary supplier of palladium. Since then, disruptions to this supply have kept prices high. As a result there has been a decrease in demand for palladium in the autocatalyst industry, and since palladium has a narrower demand base (Mineweb 2002) its future is more uncertain. The global demand for platinum fell last year by 9% at the expense of platinum, particularly in the autocatalyst market. Similarly, demand for chemical applications fell last year by 8%. Palladium in this sector is primarily used in the production of fine and specialist chemicals. Early last year palladium prices soared to over $1000 per oz which lead to an 18% decrease in demand for dental alloys. The impact was more significant in Europe where demand fell by 60%. There was a 70% slump for demand in palladium for electronic applications in 2001.

Trends in the demand for platinum are shown in Figure 2.5. The two main applications are autocatalysts and the manufacture of electrical equipment. Other applications include, in descending order of current demand, dental applications, petrochemical processing and jewellery.

Figure 2.5: Worldwide demand for palladium by application, 1976-2001 (from CPM Group, 2001)

The supply of palladium comes mainly from the Noril'sk deposit in Russia. 4.34 million oz of palladium came from Russia in 2001 compared to 2.01million oz from South African mines. North America and Zimbabwe & others supplied 850,000 and 120,000 oz respectively last year.

The other PGM that is mainly consumed by the autocatalyst market is Rhodium, supplies of which to all sectors fell last year. It is often used with platinum to form an alloy catalyst for petrol engined cars. The use of Rhodium in catalysts last year actually withstood the overall decrease in demand, due to slight increases in metal loading. This helped offset the lower production of petrol cars (Johnson Matthey plc 2002).

A small amount of Iridium is used in catalyst production, though again, the demand for it fell considerably, by 33% last year.

Supply of the rarer PGMs comes as a by-product of mining for other metals, as they are typically found in alloys of platinum and palladium. The irregular price patterns of these metals are linked primarily to a main product, and therefore slow to respond to its own demand conditions (Campbell and Minnitt 2001). Any shift in supply and demand is typically large relative to the small market and causes a large reaction in price.

As the rarer PGMs are mined as by-products of platinum or palladium mining, the impacts on the environment of these individual metals is likely to be on a similar scale to those of platinum and palladium.

There are currently no substitutes for PGMs in the manufacture of autocatalysts or hydrogen fuel cells. Various combinations of PGMs have been shown to be more efficient, but there are no examples of non-PGM metals showing the same level of performance (Financial Express 2000). Johnson Matthey has confirmed that they are not investigating alternatives to platinum for hydrogen fuel cells, and that they do not believe such alternatives would provide any advantage on a £/kW basis (Potter 2002).

A breakthrough has been made by Medis Technologies LTD in developing a fuel cell using a base metal catalyst instead of platinum. The fuel cell is a direct liquid ethanol/methanol fuel cell but nevertheless highlights the possibility of base metal catalysts for low temperature fuel cells. Another low temperature fuel cell option, the Alkaline Fuel Cell (AFC), also uses non-PGM catalysts, but the AFC has other disadvantages such as lower efficiency and the need for pure oxygen supply.

Research is presently being carried out in California to assess the potential of a possible alternative to precious metal catalysts. Biofuel cells make use of the catalytic activity of living cells and/or their enzymes. (Butlin 1999). It has been concluded through research in this area that it is feasible to genetically tailor certain enzymes for reaction on the surface of fuel cell electrodes. This represents a very long term alternative to PGMs which has yet to be proven at a significant scale.

2.7  Projections of Future Platinum Demand

We have developed a simple spreadsheet model to explore the likely impact of hydrogen fuel cell vehicles on the demand for platinum between now and 2050. The model does not provide an accurate prediction of future platinum demand, due to the simplicity of the model and uncertainties in key assumptions, but it is a useful tool for exploring sensitivities to factors such as vehicle demand growth, platinum recycling rate and the rate of penetration of fuel cell vehicles.

This section describes the assumptions and data used for the model, shows the demand projections developed for different fuel cell vehicle penetration scenarios and discusses the implications and uncertainties associated with these projections.

Assumptions and methodology

The model estimates the platinum demand from automotive applications in the years 2000, 2010, 2020, 2030, 2040 and 2050. Demand from non-automotive applications is assumed constant at 2000 levels as we have no information on likely future demand in the jewellery, chemicals, electrical and other sectors. In practice this may underestimate future demand as these applications have grown in recent years.

There is assumed to be no penetration of platinum-containing fuel cells in non-transport applications. Again this is likely to introduce an underestimate as cost targets for stationary applications are less challenging than those for transport applications. However, as discussed in Section 2.2, platinum use for fuel cell power generation is unlikely to exceed 0.1 million oz per year and so the effect is not large.

Total sales of cars worldwide are assumed to increase at a rate of 45% per decade from a 2000 figure of 48.7 million vehicles per year (WVMA 2002) to 312 million vehicles per year in 2050. The rate of growth is based on USDOE (2000) predictions of future car stock growth which indicate that there may be 3 billion cars on the road worldwide by 2050. This is consistent with projections of car demand in the period 2010-2030 in other publications (e.g. Automotive Online 2002). The above figures include light trucks and SUVs which are used interchangeably with cars in North American markets.

The model does not take account of platinum usage in vehicles other than cars and light trucks. Buses may offer an attractive initial market for fuel cells but the worldwide market for buses is very much smaller than the car market, and is growing more slowly.

Cars are assumed to have a lifetime of 10 years such that the stock is replaced after each time period and recycled platinum is returned to the market one time period after it is used in car manufacture. In reality, car lifetimes may increase due to improved materials and the faster pace of growth in developing countries where people are unlikely to be able to afford to replace their vehicles so often. Longer lifetimes would have the effect of reducing car sales and lengthening the time between platinum use in vehicle manufacture and the availability of recycled platinum.

Fuel cell stacks are assumed to last the lifetime of the vehicle. This seems a reasonable assumption since fuel cell stacks being developed for stationary applications are expected to have a lifetime of over 20,000 hours and a car is used for less than 5,000 hours over a typical life (based on 10 years x 20,000 km at 50 km/hour).

Platinum loadings for autocatalysts are assumed constant at 0.05 oz per car throughout the time period (Automotive News 2002). This is based on the assumption that emission standards will continue to become more strict, driving higher loadings, but this will be balanced by improvements in catalyst technology.

Platinum loadings for fuel cell stacks fell from about 20 oz per car in 1990 to 2 oz per car in 2000 (Cahn 2002). This is expected to fall to 1 oz per car by 2010 (Potter, 2002). Ultimately levels are expected to reach 0.2-0.3 oz per car (Evans 2002; Uravan Minerals 2002; Adcock 2000), so 0.3 oz is assumed for 2020 and 0.2 oz for 2030 and beyond.

The rate of recovery and recycling of platinum from autocatalysts is assumed to be 90% from 2010, and the rate of recovery and recycling from fuel cell stacks 95%. These figures are based on current best practice (Handley 2000) and have been confirmed in discussions with Johnson Matthey (Evans 2002). The current rate of recovery from autocatalysts is currently much lower at about 70% because autocatalysts are not always recovered from end-of-life vehicles (Evans 2002).

All cars are assumed to be either hydrogen fuel cell cars or conventional diesel or gasoline internal combustion engine (ICE) cars, i.e. there is no significant penetration of alternative fuels such as natural gas or methanol.

All new ICE cars are assumed to have autocatalysts fitted from 2010. For 2000, it is estimated that 78% of new cars have autocatalysts, based on the proportion of cars manufactured for markets in the North America, the European Union and Japan.

Demand projections

A base case and three scenarios have been modelled. The base case assumes no penetration of fuel cell vehicles while the three scenarios assume different and increasing rates of penetration, as shown in Table 2.3.

Table 2.3: Different scenarios for fuel cell penetration (% of new car sales)

Scenario 3 is probably unrealistic but represents a useful "worse case" scenario from the perspective of platinum demand. Scenario 2 is more realistic but still assumes rapid development of fuel cell technology and the widespread introduction of hydrogen fuelling infrastructure in most countries. Scenario 1 represents a slower pathway whereby fuel cell vehicles find a significant niche by 2040 for urban applications where hydrogen infrastructure is less costly to install.

The demand projections for each scenario are shown in Figures 2.6 to 2.9 overleaf. For scenarios 2 and 3 there appears to be a dip in the platinum required for jewellery in 2040 and 2030, respectively, but this is actually due to a net positive contribution from autocatalysts in those years, i.e. more platinum is returned to the market from recycling than is needed for new autocatalysts.

Figure 2.6: Base case demand projection (million oz Pt)

Figure 2.7: Scenario 1 demand projection (million oz Pt)

Figure 2.8: Scenario 2 demand projection (million oz Pt)

Figure 2.9: Scenario 3 demand projection (million oz Pt)

The base case (Figure 2.6) shows fairly flat net demand for autocatalysts up to 2020 as increasing sales of cars with autocatalysts are balanced by increasing amounts of recycled platinum entering the market from end-of-life vehicles. Thereafter demand increases with car sales such that demand for automotive applications reaches 5.9 million oz by 2050, which is equal to the total platinum demand for all applications in 2000.

Scenario 1 (Figure 2.7) shows the effect of a gradual introduction of fuel cell vehicles to reach the level of 20% of total new car sales by 2050. This pushes up total platinum demand by about 50% in 2050 compared to the base case (from 10.4 million oz to 15.1 million oz) due to increasing demand for fuel cell catalysts, which contain at least four times more platinum per car than the autocatalysts which they displace. In the longer term, the higher percentage recovery of platinum from fuel cells would help to close the gap such that a fleet of fuel cell cars would use only twice as much new platinum as a similar fleet of ICE cars if vehicle demand was constant over time.

Scenarios 2 and 3 (Figures 2.8 and 2.9) show the greater impacts of faster introduction of fuel cell cars. Platinum demand is much higher here for two reasons: the high platinum content of fuel cell cars compared to ICE cars and the 10 year lag between fuel cell car manufacture and platinum recycling from fuel cell stacks. In both scenarios, platinum demand rises to a maximum of about 25 million oz per year, which is over four times the current demand. Interestingly this peak comes earlier in Scenario 2 where the rate of penetration is slower overall. This is because this scenario assumes a large increase from 30% to 70% of total sales between 2030 and 2040, which is a greater increase in sales of fuel cell cars than over any of the time periods in Scenario 3.

Conclusions on future platinum demand and availability

The above projections, coupled with the statements from Cawthorn (1999) about accessible platinum reserves in South Africa, suggest that platinum availability should not be a constraint to the introduction of hydrogen fuel cell cars. If South Africa alone can deliver up to 5% per year additional platinum supply between 2000 and 2050, this equates to an additional 13.6 million oz in 2030, 24.8 million oz in 2040 and 42.9 million oz in 2050, which is sufficient to meet demand under any of the scenarios considered.

However there are many important assumptions and uncertainties built into this model. For example, this additional South African platinum supply would be insufficient to meet worldwide platinum demand by 2040 under Scenario 2 (realistic penetration) if any one of the following alternative assumptions is made:

• South African supply can only be increased by 4% per annum instead of 5%.
• Jewellery demand grows at more than 2% per annum - it is currently assumed to remain constant but grew by an average of 6% per annum between 1994 and 2001.
• Fuel cell stacks require more than 0.3 oz of platinum per car in 2040 - it is currently assumed that only 0.2 oz will be required but this is a factor of 10 less than current stack technology.
• The demand for cars grows by more than 55% per decade - it is currently assumed to increase by 45% per decade based on USDOE projections.

The platinum loading for fuel cell stacks is an important factor in determining the commercial viability of fuel cell cars as well as determining potential platinum demand constraints. The price of platinum is not likely to be a constraint to the introduction of fuel cell vehicles if the expected reductions in platinum loadings are achieved. At current platinum prices and the target platinum loading of 0.2 oz per car, the platinum required for a single car would cost about $90 or $1.5/kW, compared to a cost target of $50/kW for the whole fuel cell engine.

Back to top