What's New

Are Lithium Ion Batteries a Gamble?

Lithium-ion batteries have brought us a long way, and with the introduction of Tesla’s electric vehicles and its home energy storage products, lithium-ion is on its way to becoming the perceived default technology for any kind of energy storage. While there is no denying the potential of Li-ion, the unconscious acceptance of Li-ion as the de facto standard is at the least fraught with economic risk and at the most may be a dangerous concept.

Benefits of Li-ion

Let’s back up a bit. Li-ion batteries have many benefits, including the fact that they hold their charge better than other battery chemistries. In comparison to nickel metal hydride (NiMH) Li-ion loses only about 5% of its charge per month, compared to almost 20% for NiMH. They also have no memory effect, which means you do not have to discharge them completely before recharging. This is not the case with some other battery technologies. Finally, Li-ion can handle hundreds of charge-discharge cycles — a major advantage if you are using them for laptops, EVs, and other intensive-use applications.

Disadvantages of Li-ion

On the other hand, Li-ion has well-documented disadvantages in its standard form. For one, these batteries start degrading as soon as they leave the factory. They typically last for only 2 or 3 years from date of manufacture – whether you use them or not. Li-ion batteries are extremely sensitive to high temperatures and will degrade faster the hotter they get. This can be a significant problem when looking at certain kinds of operating environments. Another problem is that if you completely discharge a Li-ion battery it will be ruined, which can be a costly lesson, especially because Li-ion must have an on-board computer to manage the battery. Management is important because with current Li-ion chemistry there is a small chance that if the battery fails it will explode. We have, of course, seen news reports about unexpected explosions of these batteries on a random basis, including the so-called “hoverboards,” and we have seen the controversy surrounding the transport of Li-ion batteries on airliners (Boeing and Airbus have warned that current fire prevention technology is not adequate).

Not just safety, but failure risk

Less talked about, but with larger implications for applications like electrical infrastructure, is the mathematical and statistical failure rate. While reliability of Li-ion battery technology has improved over the past decade from an individual cell failure rate of 1 in 200,000 to 1 in 10 million, what does this mean today from a risk perspective? The good news is that “failure” is not the same as “explosion,” so the potential for direct human harm has indeed been mathematically reduced. The bad news is that not all cells are created equal — this lower failure rate is really only applicable to 18650-type cells produced by certain high-quality manufacturers, says Alex Bistrika, president of eChemion, a Corvallis, Ore.-based technology firm that maximizes battery capability.

However, let’s suppose that Li-ion continues to be viewed as a viable solution for higher volume energy storage. This means the energy needed to power EVs, households and other emerging applications such as kinetic recovery energy systems used by transit systems, and possibly grid-related storage for renewable energies or grid balancing. Is the mathematical failure rate acceptable for these industries?

Even more importantly, given varying standards of quality control for Li-ion manufacturing, does this actually increase the odds of catastrophic failure? All it takes is microscopic contamination to create a major risk. Battery education website BatteryUniversity.com warns that “protection circuits can only shield from outside abuse, such as an electrical short or faulty charger. If, however, a defect occurs within the cell, such as a contamination of microscopic metal particles, the external protection circuit has little effect and cannot arrest the reaction.” This does not bode well for batteries manufactured in regions where quality control is not a priority.

Gambling on the Odds

Taking a closer look at these mathematical failure rates and comparing them to some other odds can help bring us a better picture of the potential implications. Think of a Li-ion battery cell as a die that you can roll. When you roll a die you have a 1 in 6 odds of rolling a 2. Similarly, with a single high-quality battery cell you have a 1 in 10 million chance of failure. As you add cells, the chance of each individual cell failing is 1 in 10 million, but the odds of failure increase significantly when you have many of these individual cells — each with their own chance of failure — in a single battery or in an installation.

Suppose that you install a much-vaunted Tesla Powerwall storage battery in your home. With approximately 900-plus cells, your odds of a high-quality cell failure move from 1 in 10 million to about 1 in 11,000. To add some tangible perspective, this roughly the same odds that you will be struck by lightning once in your life. Now think about your newly purchased Tesla vehicle with more than 7000 cells. Increasing your cell count anout 7 fold, you have still only increased your failure odds to about 1 in 1,400, roughly half the odds of you dying as a pedestrian. For most, this is not enough to dissuade you from buying an innovative product with some pretty amazing benefits (especially if you are recharging your car with solar).

But let’s look at the broad interest in batteries for grid-level energy storage. Say you want to set up a 4MWh installation similar to the recent pilot project deployed by Unienergy (UET) in northern Washington State for Avista Utilities, a project that uses vanadium flow batteries, not Li-ion. If you are considering Li-ion, your installation will contain a string of nearly 600,000 cells. Your failure odds just jumped exponentially – for a mission-critical grid application – to roughly 1 in 17. This highly possible scenario gives you about the same odds as rolling two dice for a total of seven. Acceptable for a grid-level installation? Probably not — at least not without costly redundancy provisions, dividers between cells,  and class D foam-based fire suppression systems — yet we continue to see proposals to the industry for grid-level storage using Li-ion.

Looking ahead: application suitability

The solution is not restriction or banning of Li-ion, this would have a major adverse economic effect on many established and emerging industries. Rather, the likely solution lies in choosing application-appropriate use of Li-ion while rapidly fostering growth of other energy storage solutions for high-capacity storage including caching of energy from renewables. Redox flow batteries and compressed air offer one solution for the industry, with a long-term cost structure similar to Li-ion battery industry. Additionally, because these technologies last for so many more cycles, the overall cost of electricity is likely to be many times lower, say industry experts including Ramez Naam.

Of course, new solutions will be forthcoming for grid-scale use. ARPA-e and theU.S. Department of Energy have successfully fostered public-private partnerships since 2013 to significantly advance new storage technology, with a number of efforts bearing fruit. ARPA-e recently touted reaching several “holy grails” in energy storage that puts the industry on a steep j-curve growth rate. It’s an exciting time. Let’s find the right solutions for the right uses.

About Tom Breunig (203 Articles)
Tom Breunig is principal at Cleantech Concepts, a market research firm tracking R&D projects in the cleantech sector. He is a technology industry veteran and former international marketing and communications executive who has worked with organizations in semiconductor design, water monitoring, energy efficiency and environmental sensing. He has spoken at numerous technology and energy conferences.
Contact: Twitter

Leave a comment

Your email address will not be published.


This site uses Akismet to reduce spam. Learn how your comment data is processed.