$/kWh: And the lowest price is…
11 February 2011
Perspective by The Townsend Company
...Probably not the battery that you want to put on your vehicle, but that seems to be the battery that most people are trying to buy. It has unfortunately become the cliché that whenever a battery buyer approaches a battery company the first statement heard is, “Don’t tell me anything about how special your battery is, just tell me the dollar per kilowatt hour ($/kWh) price.”
Now although it’s a useful metric, if it’s taken on its own without context it becomes a distracter as lithium battery chemistries remain different and despite much effort, are not quite ready for commoditization yet. Finding a common yardstick to compare Lithium-ion batteries against is not a simplistic task and is one that leads to great frustration as no two batteries nor customer applications are the same. The $/kWh metric has unfortunately become overused as industry bodies and governments have tried to establish targets of $250/kWh to drive the battery industry towards greater competitiveness.
Most battery companies will claim they are able to achieve this. The question becomes at what level—cell, module or pack—and when will it be validated for actual commercial use and not just sampled for proof of evidence. If a battery user approaches an application and purchase with this knowledge then they will quickly be able to downselect their battery choices and formulate a value proposition for their best purchase. To quote Sy Syms from the 1980’s, “An educated consumer is our best customer,” and this remains very true today, especially in the world of Lithium-ion batteries.
There are a number of factors that must be considered when evaluating a battery’s competitiveness. These begin with an assessment of the specific chemistries suitability for the desired application—EV, PHEV or HEV [electric, plug-in hybrid electric or hybrid electric vehicle]—and its energy and power requirement. Once that is determined, a commercial comparison metric can then be established that considers the energy and/or power that is required, the specific vehicle application, desired vehicle use and targeted life requirement. $/kWh fails to do this as it’s simply a ratio of the price of the battery and the day one advertised energy.
The more appropriate metrics that should be used for commercial and performance comparison are based on the energy or power that can actually be used and how long they can be used in an application based on use and time.
Considering this proposition for the three distinct uses of batteries in the automotive market—EV, PHEV and HEV—each one needs to be considered independently as each of their applications demand a different performance attribute.
EV applications require an energy solution in order to achieve a desired mileage range, typically 100 miles. A battery with a large amount of energy ( kWh/kg or kWh/liter), characteristically is the foundation of an EV vehicle battery. However, the purchaser then needs to understand how much of the batteries’ energy can actually be used, the usable energy, as different chemistries have different state of charge (SOC) windows that can limit the nameplate or advertised batteries’ energy from 60-95 percent. Therefore, the more appropriate commercial performance metric would be “$/kWh of usable energy.”
However, there are a number of secondary factors that have significant effect on the battery, specifically its size. Firstly, the environment the battery will be subjected to temperature extremes; charging rates; and drive cycle applications. Secondly, the vehicle’s system, specifically mechanical packaging and power train voltage. Last but by no means least, the expected life of the battery. These factors combined may require the used SOC window to be further reduced to achieve the expected life as temperature and drive cycle factors are considered. This could potentially lead to an “over sizing”, or increase in the size of the battery pack, in order to obtain a life rather than an initial range goal.
When these factors are considered, $/kWh of usable energy still falls short of providing a true measure of value that includes a life perspective. Therefore, the most pertinent performance metric becomes “$/kWh of usable energy per mile.”
For example, at a cell level we may have $350/kWh, translating to a battery pack level of $500/kWh with a chemistry that has a 70 percent SOC, leading to a $710/kWh usable energy. When based over life (8 yrs. 12,000 miles a year) there is a 96k miles requirement, resulting in a $710/kWh/96k = $7.40/kWh of usable energy per mile.
PHEV applications require a combination of both energy and power, with current applications targeting approximately 40 miles of EV range followed by HEV cycling. The ratio of the desired electric range to hybrid drive will determine the most applicable commercial performance metric. With the power requirement factor of the drive cycle becoming more prominent than in a pure EV ruling out high energy low power chemistries in favor for more balanced energy to power solutions.
As the EV range will typically determine a battery’s size the $/kWh usable energy per mile commercial performance metric is still be the most appropriate, although the $/W throughput cost should also be considered.
HEV applications are power applications, rather than energy applications, transferring watts of power over a targeted drive cycle, usually compromising acceleration and regeneration events. An HEV pack will undergo hundreds of thousands of these cycles, which to a great extent, are not related to the number of miles the vehicle is driven as typically power will only be delivered during acceleration events.
Therefore, for a power battery the $/kWh usable energy per mile metric is not a relevant metric for a HEV pack. The appropriate commercial and performance metric should be a measurement of power, to a required drive cycle, targeted life and price of the battery. The appropriate commercial metric for a HEV pack should be $/MWh of throughput, where throughput is the number of cycles required to reach the batteries targeted end of life requirement.
For example: A 100 kW HEV pack is priced at $6,000. It is required to have a 120 MWh of throughput in order to reach an 8-year life and still be capable of a minimum of 40 kW of power (in a 260V system). The drive cycle, end of life power and thermal management requirement will size the pack and specifically the number of parallel strings and battery cost. Therefore the $/Mwh of throughput would be $6,000/120 MWh = $50/MWh throughput.
About The Townsend Company
The Townsend Company LLC was founded in 2010 as a consultation and business development practice to help companies wanting to enter, grow or become more profitable in the alternative energy market. Prior to founding the company, Glynne Townsend, President and CEO, led the revenue growth for A123 Systems in the automotive and grid markets and is a leading commercial authority on Lithium-ion batteries. For more information please contact: [email protected]
The reason a battery buyer might ask about a product's $/kWh first is because they know how many kWhs they will need in their vehicle and how much they can afford to spend on that amount of storage. If a product is too expensive, it can be ruled out quickly and said buyer won't have to listen to a long-winded spiel about the many wonders of a certain cell.
I think the vast majority of people who are putting many thousands of dollars into batteries know enough to ask about all the other pertinent metrics for their needs.
Also, specific gravity is more often expressed in Wh/kg rather than kWh/kg.
Posted by: drivin98 | 11 February 2011 at 06:49 AM
Do we buy vehicles by the pound? May be we do. That may be one of the reason we have so many 3+ton monsters on the roads.
Large pick-ups are much cheaper than Mercedes or BMWs by the lbs.
So, we buy more over sized pick-ups that higher quality cars.
Posted by: HarveyD | 11 February 2011 at 07:35 AM
"per mile" is not a quantity intrinsic to the battery, and is therefore not useful for comparing two batteries.
The second metric, $ per total MWh delivered over the lifetime of the battery is a more useful metric for all applications (EV, PHEV, HEV) although it makes some assumptions about what defines the battery life and those assumptions need to be both realistic and clearly stated for the metric to be useful.
Posted by: Allen | 11 February 2011 at 09:36 AM
IMO most useful would be $/useful kWh/full cycle indicator. Those consultants are making simple things complicated. It's their job. For HEV this comparison could be performed at set power level (for instance usefull battery power 60 kW).
Posted by: Darius | 11 February 2011 at 12:25 PM
P.S. It could happen lead acid bateries become best choice. For this could be avaluated additional automobile cost transporting 1 kg or liter of battery in $ terms and set kg/useful kWh indicator. That would be job for consultants.
Posted by: Darius | 11 February 2011 at 12:38 PM
$710/kWh/96k = $7.40/kWh of usable energy per mile.
By my math it's 7.4 cents / kWh usable energy per mile.
Posted by: cjmccmd | 11 February 2011 at 05:49 PM
cjmccmd,
You nailed it. Everyone always freaks out when they talk about paying $8,000 or $9,000 for a battery pack. What they don't realize is that the average person is paying between $1,500 and $2,000 a year for gas when you get up around $3.25 a gallon.
If your battery is warrantied for 10 years, then you're coming out ahead. It's the fact that you see the cost of the battery explicitly and it's up front that makes everyone crazy.
Personally, I'd much rather pay for the batteries and electricity in any way, shape or form vs. the foreign oil.
Posted by: DaveD | 12 February 2011 at 06:35 AM
$/kWh is still the number one metric because no automaker will consider a battery with a narrow usable SOC window (too heavy) or something that can't manage the lifetime cycles required (eg 1,000 deep cycles).
Posted by: clett | 12 February 2011 at 12:40 PM
This Townsend thing sounds like something to sell to innumerate managers; engineers are going to look at units like "$/kWh/mile" and say "That's not right, that's not even wrong."
Agreed that lead-acid could be a winner. I was told recently that Firefly Energy's battery failed to get DOE money for reasons that are technical rather than political, but I would not be the least bit surprised if the political requirement for 10 yr/150,000 mile warranties was the ultimate stumbling block. We'd be much better off with cheap batteries even if they only lasted 50k miles.
Posted by: Engineer-Poet | 12 February 2011 at 01:52 PM
Agreed with Allen that "$/kWh/mile" is not battery-specific but it is close to one that is. CSIRO were quoting "$/life time kWh" when comparing batteries and in particular their variant of the lead-acid battery. The point was to sum the usable energy over the life-time of cycling.
Posted by: DavidJ | 13 February 2011 at 06:14 AM
"resulting in a $710/kWh/96k = $7.40/kWh of usable energy per mile."
If the 96 was Kilo Miles in the formula, then result should be $7.40/kWh of usable energy per Kilo Mile, not by Mile...
Posted by: CARL75014 | 14 February 2011 at 11:37 AM
On my side I'll check :
1/ The right number of usable KWH capacity needed 1st
Mini is 50KWH for me, to start adopting Plug-In EVs with extended range, in order to get #200 Miles of pure EV range with full comfort, and only 2 x charges per week for my local commutes,... up to 130 KWH for #500 Miles per charge, allowing no more than 1 charge per week for local commutes, plus full week end trips coverage, plus significant parts of vacation trips covered w/o any petrol).
2/ The Number of WH/KG, WH/Litter to see if my big battery can reasonnably fit in my prefferred SUV formfactor.
3/ The number of cycles the battery will do before falling <80% capacity (For >10 years usage, I'd want > 1200 x cycles for 50KWH battery, or > 600 x cycles for 130 KWH).
4/ Finally I'll check the time required to fast recharge it practically, since with <500 Miles of EV range I'll need to keep an Extended Range ICE generator, while beyond 500M of range, I could imagine to drop that security and move to full EV, .. IF and only IF the battery could be recharged in <20mn, almost everywhere, when that becomes available.
5/ Then I'll ask for the price of the battery that could do that, since my target SUV vehicle price is in the priceband of $50K to $75K, means battery can't take more than half of that amount, assuming a real full car of the future, with only electric tracting engines, and an extended range generator optimized for constant speed operations (Micro Turbines, or Rotational engines,...etc).
Posted by: CARL75014 | 14 February 2011 at 12:05 PM
What the US needs right now is not a good battery but factories that make automobile fuel from coal and natural gas. Good enough batteries for plug-in-hybrid cars have been available for over twenty years. All the US gets is the expensive TESLA and VOLT; Both have too much horsepower as do most automobiles on the market. ..HG..
Posted by: Henry Gibson | 16 February 2011 at 04:52 AM
I want to knwow how the percentage cost of battery materials. Do Anodes cost more than cathodes, Electrolites and Separators combined. Can U please shed dum light in this regard guys.
Posted by: Spoogzx | 11 March 2011 at 05:32 AM