Tuesday, March 30, 2010

How Much Lithium Does a Battery Really Need?

EV World

Technicians transfer prototype Volt lithium-ion battery pack at GM's Global Battery Development Laboratory in Warren, Michigan.

The question of how much Lithium or Lithium Carbonate is required per kWh of battery capacity has become a matter of some importance due to the limited availability of Lithium for EV applications. Questions as to the feasibility of establishing mass production of more than a few million PHEV rechargable battery packs per year are in part met with assurances that the quantity of Lithium required per kWh is low.

A range of figures for the quantity of Lithium required per unit battery storage capacity (kWh) have been published or quoted recently. Some of these figures quote the minimum theoretical quantity of Lithium per kWh as if this is achievable in a practical device. Other figures are also unrealistically low.

To address this issue objectively, we have produced a briefing paper to illustrate for strategic planners in the automotive industry how real world battery efficiency differs from theoretical.

In a recent report to investors (“Lithium Hype or Substance” 28/10/09), Dundee Capital Markets assume a Lithium Carbonate requirement of 425 grams LCE per kWh (80 g of Lithium metal). This is equal to the theoretical minimum amount of Lithium needed per nominal kWh in a 3.3 V (LiFePO4) system. It is unrealistic to expect to approach that capability in a real battery.

In a more detailed presentation from ANL (“Lithium Ion Battery Recycling Issues”, Linda Gaines, Argonne National Laboratory, 21/5/09), estimates are presented varying between 113 g and 246 g of Lithium (600 g and 1.3 kg LCE) per kWh for various cathode types of batteries all with a graphite anode; a Lithium titanate spinel anode battery is shown as having a high requirement of 423 g Li (2.2 kg LCE) per kWh.

This range of figures illustrates the difficulty that may exist in modelling LCE requirements for strategic planners. The briefing paper describes the main factors that occur in a real battery to reduce its effective capacity and estimates a realistic figure for the quantity of raw LCE that should be assumed to be required per kWh battery capacity.

In a real battery, the main factors which reduce capacity below the theoretical maximum are:

Irreversible capacity loss: when the battery is first charged, some of the Lithium becomes bound up in the anode and cathode and electrochemically inactive. This can be as high as 50% of the Lithium originally put into the cathode before the battery is charged for the first time. In particular Lithium forms a layer known as a Solid Electrolyte Interphase on the anode which increases the internal resistance in the battery and internal energy losses.

Discharge rate: this is the major variable which reduces day to day effective capacity while the battery is in use. The Energy batteries required for PH(EV) use are more sensitive to this than power batteries and the problem is further exacerbated by using small batteries in a PHEV. Up to 50% of the effective capacity could be lost at medium to high speeds. Manufacturer capacity figures that only apply at low discharge rates are of little use in determining a realistic benchmark for PHEV battery capacity, for which capacity at the 1C rate at least should be used for a realistic indication.

Cycle life capacity fade: Batteries lose capacity as they are cycled, particularly if deep discharged (charge depleting mode) as will be the case in a PHEV. EV batteries will be 25% larger than the nominal or useable stated capacity to allow for capacity fade to 80% of the initial capacity at end of life, without the driver experiencing any perceived reduction in range as times goes on. Therefore 25% more Lithium will be required per kWh than the nominal rated capacity.

Electrochemical factors which reduce the theoretical capacity of a battery include polarisation, internal resistance, electrolyte conductivity, separator conductivity, cation transport number, cation activity coefficient and order/disorder and particle size within the electrodes. Discharge rate is the main operational factor which reduces energy capacity because theoretical capacity only applies at zero current: as soon as current is drawn from a cell it cannot but lose “free energy” (delta G) and capacity will fall. The discharge rates for a PHEV battery will be high, at least 1C on average.

The capacity of Lithium Ion batteries is also cathode limited, i.e. limited by the ability of the cathode to accept Lithium ions from the anode when it is discharged. For the LiFePO4 cathode material the capacity is often quoted as 170 mAh/g. However this is the capacity at a very low discharge rate and often for a power battery with thin electrodes and low total capacity. For a PHEV energy battery using thicker electrodes but still a high relative discharge rate, cathode capacity of the LiFePO4 material could fall to 90-100 mAh/g at high speeds – in other words, available capacity falls below 60%. Given that PHEV batteries will still be relatively small (16 kWh maximum, many will be less) and the nominal capacity may be quoted at an unrealistically low discharge rate, realistic capacity at say the 1C rate will be 35% lower than the headline figure. This means a significant percentage of the Lithium in the anode cannot be used since it cannot be accepted by the cathode at a real discharge rate and this will increase the effective amount of Lithium required per actual delivered kWh.

If we look at the theoretical specific energy of a LiIon battery, the figures widely quoted are between 400 and 450 Wh/kg. The actual specific energy achieved is between 70 and 120 Wh/kg. Therefore practical LiIon batteries are using some four times as much Lithium per kWh as the “theoretical” quantity or more. This translates into some 320 g of Lithium or 1.7 kg of Lithium Carbonate per kWh.

If we then add 25% to that figure to allow for cycle related capacity fade, 400 g of Lithium will be required. Then allowance must be made for losses in purifying raw technical grade Li2CO3 into low-sodium battery grade material: a yield of 70% increases the raw LCE requirement to 3 kg per kWh.

Therefore 3 kg of raw technical grade Lithium Carbonate will be required per kWh of final usable battery capacity.

At 3 kg raw technical grade LCE per kWh, current global production of some 100,000 tonnes raw LCE would be sufficient, if available, for some 2 million 16 kWh batteries per year. Even at an optimistic 2 kg LCE per kWh assuming very high purity yields, production would be sufficient for only 3 million 16 kWh PHEV batteries per year.