Energy Storage News

FZ Sonick Plans to Manufacture Molten Salt Sodium Nickel Chloride Batteries for Mobile and Stationary Electricity Storage Applications

A sister technology to NGK Insulators molten salt sodium sulfur (NaS)batteries is molten salt sodium nickel chloride batteries also called Zebra batteries after the trade name of their sole manufacturer, a Swiss company called MES-DEA. Like NaS batteries zebra batteries have a molten sodium anode separated from the cathode by a tubular solid electrolyte made out of beta-alumina. The cathode is primarily composed of nickel chloride (NiCl2) rather than of sulphur. During the discharge process sodium ions are formed in that anode and migrate through the solid electrolyte into the cathode. The primarily chemical reaction in the cathode is the exchange of Cl between NiCl2 and the sodium ions resulting in the growth of large crystals of NaCl. During discharge this process is reversed and the NaCl crystals decompose and the sodium ions migrate back into the anode.

Recently MES-DEA has been acquired by the Italian Company FIAMM and they have formed a new venture called FZ Sonick. MES-DEA has been manufacturing Zebra batteries for EV applications. FZ Sonick plans to add a line of batteries for stationary storage applications.

A data sheet for the Zebra EV batteries may be found here. I list some interesting numbers from this data sheet below:

I assume that BMI means Battery Management Interface.

Elsewhere is the data sheet I find this comment about depth of discharge (DOD):

System design recommendation:
- Max. degree of discharge: 80%

A data sheet for backup a battery for telecom application can be found here. Some relevant data is listed below:

Another comment in this same data sheets indicates the cycle life of the battery:

Insensitivity to the operational and environmental conditions, i. e. constant performances and lifetime over a wide temperature range -20°C +60°C, as well as with heavy cycling (>2000 cycles)

So how do these numbers compare to the same numbers for NaS batteries? According to this paper the energy density of NaS batteries is:

118Wh/Kg
160Wh/liter

This presentation published by Akihiro Bito of NGK Insulators states the cycle life of NaS batteries as follows:

2500 cycles at 100% DOD
4500 cycles at 90% DOD
6500 cycles at 65% DOD

DOD means depth of discharge.

We see that the energy density of the Zebra EV battery is very close to that of NaS batteries. The energy density of telecom backup battery is smaller, but this battery has much smaller total capacity (4KWh compared to 21.2KWh) so perhaps the penalty of the peripheral parts of the battery is higher. The cycle life is stated be >2000. If this cycle life applies at 100% DOD then it is comparable to NaS batteries. If it applies at 80% DOD then it is greatly inferior to NaS batteries. The following quotation taken from a 2003 paper by MES-DEA researchers Galloway and Dustmannn the following statement indicates that the cycle life is probably only half that of NaS batteries:

ZEBRA Battery technology has proven calendar life of more that 10 years and cycle life of 1000 to 2500 nameplate cycles dependent on operating parameters.

This statment implies that the lower cycle life number is correct which is a significant disadvantage with respect to lifetime storage costs.

Nowhere in the MES-DEA/Fiamm published material can I find a reference to storage efficiency. However, in this paper published by researchers at Centro Elettrotecnico Sperimentale Italiono (CESI) they comment about the AC efficiency of Zebra batteries under real world test conditions:

Charge/discharge efficiency, disregarding peripheral energy consumption, is about 92%, if discharge current is 20A, and is reduced to about 75% if discharge current is about 120A. The efficiency is influenced by the losses due to the system peripheral units, e.g. heating/cooling the battery, whose value depends of the intensity of the charge/discharge processes and ambient temperature, and is in general very low because of thermal energy produced during discharge and the effectiveness of the thermal insulation of the battery.

It should be noted that the battery they where testing at 32Amp-hours of storage capacity so that 20A represents and hour and a half of discharge. 90% AC efficiency is an improvement on the quoted 75% AC efficiency of NaS batteries. If battery round trip efficiency η, then an efficiency cost exist which is given by:

Efficiency Cost = [(1-η)/&eta]×Generation Cost

For NaS batteries multiplication factor is .33 and for Zebra batteries the factor is 0.11. So if the generation cost is 10¢/KWh then efficiency costs would be 3.3¢ and 1.1¢ respectively. These efficiency costs have to be added to the cost of the storage system itself. In a previous post on NaS batteries I estimated the cost of delivering 1KWh from these batteries to be 8¢ plus interest on the capital assuming a cycle life of 4500 cycles. Unless Zebra batteries can be manufactured for much lower cost than NaS batteries it seem likely that the negative effect of lower cycle life will outweigh the efficiency advantage.

It is not clear to me why zebra batteries should be any cheaper to manufacture than NaS batteries. Both batteries use solid beta-alumina electrolyte tubes and operate at high temperature in order to insure that the sodium anode remains in a liquid state. The use of nickel in the cathode rather that sulphur has been raised as a concern about the scalability of this technology. In the paper by Galloway and Dustman referenced earlier they state that the nickel cost/KWh of battery capacity is US $17.75. If you assume a lifetime of 2000 cycles at 80% DOD the cost per KWh of delivered electricity attributable to nickel is 1.1¢. This cost is significant but not necessarily a show stopper. The question is what would happen to nickel prices of this technology became an important part of a global smart grid.

One advantage which F.Z. Sonick has over other manufacturers with newer battery chemistries is the track record of Zebra batteries in the EV market. Electrochemical energy storage systems required a large up front investment which will not be recovered until after years of reliable operation. You need to convince customers that your product will work as advertised so that a track record of real world performance is valuable.

F.Z. Sonick may find a niche market in telecom backup which they are clearly targeting, but I remain doubtful that Zebra batteries can compete with NaS batteries in the utility scale electricity storage market.

Jun 16, 2011

Energy Storage News

rogerkb [at] energystoragenews [dot] com