ORNL Solid-State Battery Test: 90% Of Original Capacity After 10,000 Cycles

NOV 6 2014 BY MARK KANE 17

Voltage profiles of the (a) LiNi0.5Mn1.5O4 solid-state lithium battery discharged at different rates. The battery was charged at C/10 before each discharge measurement. (Image: Green Car Congress)

Voltage profiles of the (a) LiNi0.5Mn1.5O4 solid-state lithium battery discharged at different rates. The battery was charged at C/10 before each discharge measurement. (Image: Green Car Congress)

Researchers from Oak Ridge National Laboratory recently presented a high-voltage (5V), long-lasting solid-state battery.

According to ORNL, this new cell can achieve an extremely long cycle life of over 10,000 cycles, while retaining more than 90% of its original capacity. Additionally, the Coulombic efficiency is near 100% on the first graph.

Is this the long awaited breakthrough?

“A high-voltage (5V) solid state battery has been demonstrated to have an extremely long cycle life of over 10,000 cycles. For a given size of battery, the energy stored in a battery is proportional to its voltage. Conventional lithium-ion batteries use organic liquid electrolytes that have a maximum operating voltage of 4.3 V. Operating a battery above this limit causes short cycle life and serious safety concerns.”

“This work demonstrates that replacing the conventional liquid electrolyte with a ceramic solid electrolyte of lithium phosphorus oxynitride (Lipon) eliminates the limit of conventional lithium-ion batteries. A model battery of LiNi0.5Mn1.5O4/Lipon/Li has been operated over 10,000 cycles at a charge voltage to 5.1V.  The solid state battery retains more than 90% of its original capacity after 10,000 cycles. Such a battery has a cycling lifetime of more than 27 years with a daily charge/discharge cycle, exceeding the lifetime of most devices and even vehicles. This work infuses new life into the existing chemistry of high-voltage lithium batteries.”

Well…maybe, but we don’t see a few important bits of data, which concerns us a little. The energy density value is missing and, on the second graph, we see that efficiency is falling down together with discharge rate.

At 2C rate (30-minute discharge), you can take just half the energy as at C/10 (10-hour discharge), which is not preferable. Old fashion liquid electrolyte lithium-ion batteries are almost as good at higher currents as at low, and we need this to be true of these batteries, especially those in electric cars. 2C is just about 50 kW from ~24 kWh battery pack in Nissan LEAF (which has an 80 kW motor).

Without maintaining efficiency at higher currents, the battery will waste all of the additional energy (if in fact energy density is higher).

Source: ORNL via Gree Car Congress

Categories: Battery Tech

Tags: ,

Leave a Reply

17 Comments on "ORNL Solid-State Battery Test: 90% Of Original Capacity After 10,000 Cycles"

newest oldest most voted

Nice lifetime.

It appears they are replacing the electrolyte, which may be compatible with many chemistries. With the voltage improvement it could lead to a slightly higher energy density.

Still completely unknown are cost and performance at different temperatures.

This in not ready to be plopped into a car. It’s still in the research phase, and what will come of it is unknown. We need to show some patience.

Tesla Model S battery cycle life is about 5000 and BYD e6 has cycle life over 10 000. However this does not mean that the life expectancy of Tesla battery would be 100 years. Real world is harsh place for batteries and lab tests tell very little about real life. Therefore real life EV battery lifetime expectation is about ten years.

This is again the same old battery bull shit that we can read weekly.

Tesla has a 8 year battery warranty, which means they only expect a few percent of them to go bad by then (or else the expense will be huge). You’re not going to get batteries dropping like flies in the next two years.

Laptops and phones don’t have 8 year warranties on their batteries. There, manufacturers want to squeeze out every minute of battery life they can so that they look good in reviews, and only care that it lasts ~2 years. They don’t have liquid cooling, range charges, etc.

Dr. Kenneth Noisewater

90Ah/kg @ 5V for 10C discharging? That would be pretty darned good.. Unless I’m reading that wrong :p

That might be per kg of active material, not total cell weight.

Ya, I’m not sure. They did the same graph with a conventional liquid battery and it was around 120 Ah/kg. So not sure what they are measuring.


That measurement is probably for the cathode, seeing how they write “vs Li/Li+”.

But Satki3 claims that their solid state battery has achieved 1100 Wh/L (so probably 400+ Wh/kg). That’s really impressive. They’re working towards mass production, though it’s most definitely going to find itself in phones and laptops first at $1000+/kWh. Phones can take that cost easily.

A battery with a higher energy density could be married with a battery with higher power, or perhaps an ultracapacitor to give over all great performance. The high power battery would act like a load leveler for the system.

Yeah, since it seems like voltage really sags on this battery with load, it would be best paired with an ultracap that is filled slowly from the battery and enables the high instantaneous current for acceleration.
And both components being solid state, should be long lasting and reliable.

And let’s throw some graphene in there. Because, why not. It can apparently do anything, with the 9000 breakthroughs that are supposed to come of it.

Wow, that’s amazing. But how much does it cost?

95% capacity after 10,000 cycles is pretty good. Would work great in a hybrid. without knowing the cost and energy density not sure it would be a best fit for a BEV.

Exactly! Imagine a 85 KWh, 265 mi EPA Tesla with 10,000 cycles = 2.65 MILLION miles. What we really need are cheaper, lighter, more dense batteries, maybe 2000 cycles. Even a 150 mi 50 KWh battery = 300,000 mi. That’s plenty of miles before you need to replace the car (except maybe taxis).

From having followed EVs and battery tech from the sidelines, this ONL paper seems to be validating the approach Sakti3 has been taking.

“At 2C rate (30-minute discharge), you can take just half the energy as at C/10 (10-hour discharge)”

Maybe I’m looking at the wrong graph or reading it incorrectly, but what I see is that at C/10 you hit the knee of the voltage curve after about 100mAh with an average voltage of around 4.6. The 2C case, you hit the knee around 85mAh with an average voltage of about 4.3.
So 4.3v*85mAh / 4.6v*100mAh = 366Wh / 460Wh = 79.5%.

The C/2 case would be 4.5v*95mAh which is about 93% of the C/10. That would be the highway driving speed for 2 hours use case for a 45kWh-60kWh pack.

Seriously. Where does this “just half the energy” assertion come from. Certainly not from the discharge curves graph. My back-of-the-napkin comes up with ~80% of the energy for the C2 curve, just about the same as Sublime’s calculations.

Mark, I think you’ve got a correction to make here! (Did you misread the y-axis as starting at 0V?)

I believe that energy is the area under the curve.

You did a wrong calculation. Energy is the area under the curve, but look at the curve, the starting voltage is not zero. You need to learn some high-school physics 🙂