Rechargeable Lithium ion batteries have high energy density and high power density as well as pretty good cycle life (>1000 cycles at 100% depth of discharge) and good round trip efficiency. Historically lithium on batteries have been relatively expensive but have found a performance niche in consumer electronics because of the willingness of customers to pay a price premium for long run times. They are now finding a performance niche in battery electric vehicles for the same reason.
Recently, however, I have seen a number of announcements on the internet about companies planning to produce grid energy storage systems using lithium ion batteries. This trend seems somewhat surprising given that cost is much more critical in large scale grid application than in consumer electronics. So what is the current cost of lithium ion batteries? I have found it difficult to get direct (as opposed to hearsay) evidence about battery prices. I recently made the following unscientific survey of battery prices. I found a web site called DealExtreme which offers 18650 lithium ion batteries from a number reputable manufacturers. The 18650 format is the format used by Tesla Motors for the Model S BEV. Spot checks of the DealExtreme prices against a few other web sites show them to have equivalent or lower prices. Looking through the various offers on DealExtreme the lowest prices/kWh (typically on orders of 40 or more batteries) for each manufacturer that I found are as follows:
Averaging all four prices we get $508/kWh. I have seen cost numbers (as opposed to prices) quoted at more than factor of two lower than the above number. It is possible, of course, that the battery manufacturers are colluding against consumers and making huge profits on sales of small quantities of batteries. But for the purposes of this article I am going to depend on actual market place offers that I have seen with my own eyes rather than on hearsay evidence. This article by The Economist claims that battery manufacturers are notoriously secretive about costs but estimates current prices in the range of $400/kWh to $750kWh which is consistent with my informal survey given above.
To estimate the cost of electrical energy delivered from lithium ion batteries we need to know the expected cycle life. The batteries listed above claim a cycle life of greater than 1000, but how much greater? I will assume the answer is not greater than 2000 or the battery companies would claim this number as a bragging point. So I will pick an arbitrary number of 2000 cycles with an average capacity of 90% of the rated capacity of the new battery. I will also assume a round trip AC to AC storage efficiency of 90%. In this case battery contribution to energy storage costs (ignoring interest) for the moment is:
cost = $508/(2000*.9)=$.28/kWh
This cost is already high and does not include balance of system costs (i.e. packaging into a large container, battery management system, the inverter to convert back and forth between AC and DC current, land rent/taxes, energy loss during the round trip storage cycle, etc.). Furthermore if one does standard discounted cash flow analysis one must assume that money not invested in an energy storage system could be put in index funds and increase in value at a rate of 8% per year for all eternity, and one must therefore count this lost income as a cost in investing in energy storage infrastructure. I believe that in the fullness of time this kind of a assumption will be shown to be nonsense, but if you are hoping to shore up business as usual financial capitalism with renewable energy generation plus storage this assumption is necessary.
The battery chemistry with the largest installed based for long period (i.e. several hours) energy storage is sodium sulfur batteries (the sole manufacturer is NGK Insulators). A while back I estimated the energy storage cost from the published price of a 1800MWhr system which was to be purchased by UAE (I am not sure whether the sale was ever completed). The number I came up with was $0.16/kWhr + interest. This cost is half of that quoted above largely because of longer cycle life (4500 cycles at 80% depth of discharge), but there is no sign that NAS batteries are revolutionizing battery based grid storage.
Of course energy storage can play a variety of different roles in supporting the electricity grid. Load shifting applications in which several hours worth of power generation are stored and return to the grid at a different time of day are among the most expensive for battery based energy storage. The amount of storage required is high and the frequency of usage is low (i.e. once per day or less). If you try to lower costs by very long cycle life you are faced with long payback times (i.e. 5000cycles>=14 years). For grid support services with shorter and more frequent cycles high power lithium ion batteries may be able to compete more easily. Not as much storage is needed and if multiple cycles per day are required the payback times are decreased.
Startup company Light Sail Energy is trying to get into peaking power markets using isothermal compressed air energy storage. They have published a graph in which they compare the levelized cost of various conventional power sources to the projected costs of their compressed air system. The graph shows gas peaker plants range in cost from $0.12/kWh to $0.36/kWh. I believe that the high end of this range represents plants that support very short time period frequency regulation which have very low capacity factors. For gas peaker plants the capital cost is the same regardless of whether the plant is supplying power for 40 seconds at a time or for 4 hours at a time. For lithium ion batteries hours long storage periods require the manufacture of a large volumes of expensive complex electrode materials. Furthermore long time period grid support is likely to be required once per day or less implying that the cost savings of a long cycle life may take many years to be realized. For shorter time period voltage support with multiple cycles per day the value of the batteries could be realized over a shorter time period. Therefore these short time period services may be a niche in which lithium ion batteries can compete.
Of course my price estimate give above is a current price. The question arises as to how rapidly Li-ion battery prices are descending. I personally do not have access to any data which allows me to answer this question. One negative sign about the trend of Li-ion battery prices is the decision of electric car maker Tesla Motor to get into the battery manufacturing business by investing 4 to 5 billion dollars in a Li-ion battery giga factory. The factory should be completed by 2020 and is supposed to bring down the cost of Li-ion batteries by 30%. Tesla motors original strategy was to avoid batter manufacturing and simply ride the cost/performance curve of batteries for consumer electronics which already have a large customer base which is hungry for improved performance. Tesla's decision to reverse this strategy and make huge (and widely regarded as risky) investment in battery manufacturing is an indication that they are not happy about the cost reduction curve of lithium ion batteries.
And, by the way, a price reduction of 30% from the number given above is not going to initiate an energy storage revolution, although it might allow lithium ion batteries to compete in the market for short time period/frequent cycle voltage support applications.
Younicos is an energy storage service company (They do not manufacture batteries) with the amusing and optimistic company slogan: Let the fossils rest in peace. They are proposing to use multiple battery technologies in their grid storage projects: Li-ion batteries for high power, short time period applications and Sodium-Sulphur or Vanadium Redox Flow batteries for long time period applications. I have already mentioned that NGK Insulators Sodium-Sulphur batteries have been on the market for more than a decade and they have not yet set the energy storage market on fire in spite of cost that are lower than lithium ion batteries.
My current judgment is that multiple hour storage applications utilizing lithium ion batteries are not on the verge of economic competitiveness in the near to intermediate term.
Oct 4, 2014Energy Storage News
rogerkb [at] energystoragenews [dot] com