Latest / eMotors: Electric Revolution / Electric Vehicle Battery Longevity and Degradation Analysis
Transcript
- 0:09Welcome to the Deep Dive, the place where we synthesize expert
- 0:13research and real world data to deliver the essential knowledge
- 0:17you need. Today we are diving deep into,
- 0:20well, probably the biggest source of anxiety surrounding
- 0:22electric vehicles, the battery. You hear it all the time, right?
- 0:26Especially if you follow the EV conversation, particularly from
- 0:30folks who are maybe a bit skeptical, they often voice that
- 0:33common, slightly chilling concern.
- 0:35Oh, it won't last more than a couple of years.
- 0:37And replacing it, that will cost a fortune.
- 0:39And you know, that fear isn't totally out of nowhere.
- 0:42It's kind of rooted in our everyday experience, isn't it?
- 0:45Like with old smartphones or laptops where the battery just
- 0:47gives up? So people look at that big
- 0:49expensive pack under the car floor and see maybe a ticking
- 0:53time bomb. Our mission today is pretty
- 0:54simple. We're going to use some really
- 0:55advanced scientific modeling plus vast real world data.
- 0:59We're talking 10s of thousands of vehicles here to try and
- 1:02resolve that anxiety. But crucially, we want to go
- 1:04beyond just saying don't worry, we'll reveal actionable ways you
- 1:08can actually manage battery's lifespan and maybe even more
- 1:11fascinating, show you how car engineers have designed some
- 1:15pretty sophisticated control strategies to boost durability.
- 1:19OK, let's unpack this. Yeah, and what's fascinating
- 1:22right off the bat is how the high level data really clashes
- 1:24with that popular anxiety. But the full picture is
- 1:28definitely more nuanced. Our source material today is
- 1:30actually quite nicely split. On one hand we've got large
- 1:33scale empirical, you know, real world data from organizations
- 1:36like Geotab. This shows exactly how
- 1:38degradation happens out there with everyday drivers.
- 1:41And then on the other side, we have these advanced scientific
- 1:44models. They're often called pseudo 2
- 1:45dimensional or P2D models. Engineers use these to really
- 1:49understand the microphysics, the chemistry of aging, and then
- 1:52design the control systems in the car.
- 1:55The core tension really running through all this research is
- 1:58this fundamental trade off to maximize battery life.
- 2:01You know, to get that potential two decade lifespan, we hear
- 2:04about engineers and by extension you as a driver often have to
- 2:07compromise a bit on vehicle performance.
- 2:09Compromise how? Like slower acceleration.
- 2:12Exactly. It could mean potentially
- 2:14limiting maximum acceleration, maybe restricting the usable
- 2:17driving range slightly, or even extending charge times.
- 2:20It's the central balancing act. So our goal for you the
- 2:24listener, is to walk away understanding the key control
- 2:28variables, what you can manage through your driving and
- 2:31charging habits, and what the vehicles sophisticated battery
- 2:34management system the BMS is actually handling behind the
- 2:38scenes. OK, let's start then with the
- 2:40big picture, the real world data, because honestly, the news
- 2:43here is pretty positive, much more optimistic than the
- 2:45skeptics might lead you to believe.
- 2:47We got this recent data from Geotab analyzing what was it,
- 2:51over 10,000 electric vehicles in actual use.
- 2:54And the results? They're incredibly encouraging.
- 2:57They really are. The average annual degradation
- 2:59rate for EV batteries isn't just holding steady, it's actually
- 3:01improving quite dramatically. Just a few years back and
- 3:04Alice's put it around 2.3% per year.
- 3:07That was a 2 or 19 figure. Yeah, right.
- 3:09But the newest analysis, it reveals that rate is now down to
- 3:12an average of just 1.8% per year.
- 3:15That's well, that's significant improvement in durability in a
- 3:17pretty short time. It is significant and that
- 3:19reduction, it speaks volumes. It's not just about small tweaks
- 3:23in battery chemistry, though that's part of it.
- 3:25Critically, it's about the maturity and frankly, the
- 3:28sophistication of the thermal management systems and the
- 3:31software controls that car makers are implementing now.
- 3:34They're learning fast. They are learning very quickly
- 3:36how to protect those cells better.
- 3:37So if this improved 1.8% annual rate holds steady, and the data
- 3:43suggests it probably will for many vehicles, the optimistic
- 3:46projection gets really interesting.
- 3:48We're talking battery packs that could last 20 years, maybe even
- 3:52more. 20 years. Think about that.
- 3:54Yeah, in places like Europe, the average lifespan of a car on the
- 3:57road right now is what may be thirteen years.
- 4:00So this means the battery is potentially said to outlast the
- 4:03car itself. It's a fundamental shift in
- 4:05thinking and that longevity it translates directly into
- 4:09retaining usable range over time.
- 4:12The research projects that even after two full decades of
- 4:15ownership, the average owner can likely expect to retain
- 4:18somewhere around 64% of the original available mileage.
- 4:21OK, 64%. So that's definitely a loss of
- 4:23range. You can't ignore that.
- 4:25But it still means the vehicle remains completely usable for
- 4:29most daily driving, you know, commutes and errands for its
- 4:32entire expected life. Absolutely.
- 4:34For the vast majority of trips, it's still perfectly adequate.
- 4:37And what about that nightmare scenario?
- 4:39The catastrophic failure, the battery just dying and needing a
- 4:43hugely expensive replacement. Yeah, that fear, the multi $1000
- 4:47paperweight idea, it's largely unfounded.
- 4:50Statistically in modern EVs, the actual failure rate where a
- 4:54battery truly dies prematurely and needs a complete replacement
- 4:57outside of the warranty period, it's minimal.
- 5:00We're talking under .5% really quite low.
- 5:02Less than half a percent. Exactly.
- 5:04So the useful mindset shift here I think should be moving away
- 5:08from worrying will the battery just fail completely and
- 5:11thinking more like, OK, how do I actively slow down the
- 5:13inevitable natural capacity loss that will happen over time?
- 5:17That's a really key distinction. It's about management, not just
- 5:20fear of failure. Now, you mentioned capacity loss
- 5:23isn't just a smooth straight line downwards over those 20
- 5:26years. Can you elaborate on that?
- 5:29The nonlinear nature of it? Right.
- 5:31It's crucial to understand the shape of the degradation curve.
- 5:34Capacity loss doesn't happen smoothly or linearly, even
- 5:37though that 1.8% average sounds low when spread over 2 decades.
- 5:41If you were to plot the battery's health, it's state of
- 5:43health, or SOH over its lifetime, you'd see a distinct
- 5:47pattern. Typically there is an initial,
- 5:49somewhat steeper drop in capacity that happens in the
- 5:52first few years of battery's life.
- 5:53The. First year or two?
- 5:54Usually, yes, maybe the first one to three years.
- 5:57Then this initial drop is followed by a long moderate
- 6:00period where the rate of loss actually slows down quite
- 6:03considerably. Degradation becomes very
- 6:05gradual, almost linear, but much flatter then potentially right
- 6:09towards the very end of the battery's usable life, and this
- 6:12could be after many, many years, you might see a final, more
- 6:16significant drop in the state of health.
- 6:18It's often signals it's approaching its retirement
- 6:20point, maybe around the 70% or 60% SOH mark, depending on the
- 6:24manufacturer's definition. OK, so let me make sure I've got
- 6:27this. If I buy a brand new EV today
- 6:29and after maybe say 12 or 18 months, I noticed my maximum
- 6:34range seems a bit lower than when it was fresh off the lot.
- 6:37That steeper initial dip is actually expected.
- 6:40It's not automatically a sign that my battery is faulty or
- 6:42dying fast. Precisely that initial settling
- 6:44in period, that steeper drop, is a known part of the battery's
- 6:48stabilization process. We can actually connect that
- 6:50back to the microphysics. The chemistry change is
- 6:52happening inside later on. But yes, knowing this curve
- 6:56exists is really essential for setting accurate expectations.
- 6:59As an owner, you don't want to panic unnecessarily.
- 7:02Good to know. So what causes this degradation?
- 7:05You mentioned 2 fundamental types of aging.
- 7:07That's right. The research breaks down aging
- 7:09into two main categories that help explain why capacity is
- 7:12lost over time. The 1st is calendar aging.
- 7:16This is degradation that happens simply because the battery
- 7:19exists, it's storing energy, and time is passing.
- 7:23It happens even if the vehicle is just parked for months, not
- 7:25being driven at all. Like rust on a parked car, but
- 7:28chemical. Kind of, yeah.
- 7:30Think of it as the inevitable slow chemical side reactions
- 7:33occurring inside the cell over time, just associated with the
- 7:36battery storing energy at a certain voltage and temperature.
- 7:39So. Even if my EV is sitting nicely
- 7:41in the garage, the aging clock is still ticking chemically.
- 7:44That's right. And the second category is cycle
- 7:46aging. This is the degradation that's
- 7:49directly related to the charge and discharge cycles.
- 7:51How much you use the battery and crucially, cycle aging is
- 7:55strongly dependent on how the battery is utilized, how much
- 7:58energy you pull out in one go, how fast you pull it out like
- 8:01during heart acceleration, and how frequently you cycle it
- 8:04between high and low states of charge.
- 8:06So calendar aging is about time and storage conditions.
- 8:10Cycle aging is about usage patterns.
- 8:12Exactly. And regardless of which type of
- 8:15aging is dominant in a particular situation, the
- 8:18macroscopic effects what you, the driver, actually experience
- 8:22are the same. First, you get a loss of total
- 8:24storage capacity. That means you're driving range
- 8:26decreases. Second, you see an increase in
- 8:29the battery's internal resistance.
- 8:31OK, what is increased resistance mean practically?
- 8:33It means the battery can't deliver power as efficiently or
- 8:36accept charge as quickly. So you might notice slightly
- 8:40less punchy acceleration over time, or perhaps reduced
- 8:43effectiveness of regenerative braking, especially when the
- 8:46battery is cold or very full. Got it.
- 8:48Capacity loss and power fade. OK, let's dive into what causes
- 8:53this aging. What are the main stress
- 8:54factors? OK, so now we drill down.
- 8:57Decades of empirical studies on lithium ion batteries have
- 9:00really identified 3 main stress factors that accelerate both
- 9:03calendar and cycle aging. The first one, and probably the
- 9:07most easily understood, is temperature.
- 9:10Let's call it T Cycling a cell, or even just storing it at high
- 9:14temperatures degrades it significantly faster.
- 9:16Period. And the real world data backs
- 9:18this up strongly. Absolutely.
- 9:20The Geotab data confirms it clearly EVs that operate
- 9:24primarily in hot climates, and the research often defines this
- 9:27as places experiencing, say, more than five days per year
- 9:29with the temperature gets over 80° or about 27° C Those
- 9:34vehicles show a statistically significant accelerated decline
- 9:37in battery health compared to identical vehicles driven in
- 9:40temperate climates. So just to be clear, the
- 9:42lifespan of my EV driven mostly in Phoenix, AZ is likely going
- 9:46to be shorter than the exact same model driven mostly in,
- 9:49say, Oslo, Norway. That is a very accurate
- 9:52generalization based on all the available data.
- 9:54It's one of the most consistent findings.
- 9:56Which immediately brings us back to the car's design, right?
- 9:58Specifically, the cooling system.
- 10:00Precisely. And we have a really powerful
- 10:02specific example right from the source that illustrates this
- 10:04perfectly. Let's compare 2 cars from the
- 10:07same model year 2015. First, the 2015 Tesla Model S,
- 10:12It used a robust active liquid cooling system, coolant
- 10:16circulating around the battery modules.
- 10:18Its average annual degradation rate was measured at 2.3%.
- 10:22OK, 2.3%. Now compare that to the 2015
- 10:25Nissan Leaf from the same era. The Leaf used a simpler passive
- 10:29air cooling system, basically just airflow over the battery
- 10:32pack. It's measured degradation rate
- 10:344.2% per year. Wait, 4.2% versus 2.3%?
- 10:38That's nearly double the rate of capacity loss.
- 10:41Yeah, just because of how the car manages heat.
- 10:43Exactly. That's almost twice as fast, and
- 10:45that magnitude of difference really hammers home why thermal
- 10:47management is arguably the single most important aspect of
- 10:50battery longevity that the manufacturer controls through
- 10:53design. So liquid cooling systems are
- 10:55just demonstrably better for making batteries last longer.
- 10:57Yes, because they actively pull heat away from the cells,
- 11:00especially during stressful situations like fast charging or
- 11:04demanding driving, minimizing the time the cells spend at
- 11:07those damaging high temperatures.
- 11:09OK, temperature is critical factor one.
- 11:11What's #2. The second major factor is the
- 11:13state of charge, usually abbreviated as SoC, or you can
- 11:17think of it in terms of the batteries voltage as they're
- 11:19directly related. Lithium ion cells, chemically
- 11:23speaking, are most stressed when they operate at extreme SoC
- 11:26levels. That means very low near empty,
- 11:29or very high near full. A lot of previous research
- 11:33established pretty clearly that operating below about 20% SoC
- 11:37and above about 80% SoC generally accelerates the
- 11:40chemical degradation mechanisms inside the cell.
- 11:44OK. And this connects directly to
- 11:45that common piece of advice you always hear for EV owners,
- 11:48right? For the typical NMC or NCM
- 11:50batteries, that's nickel, manganese, cobalt chemistries
- 11:52which are very common. The most common in long range
- 11:55EVs, yeah. The best practice for daily
- 11:57driving is try to keep the state of charge roughly between 20%
- 12:01and 80%. Don't regularly drain it way
- 12:03down, and don't constantly charge it right up to 100% if
- 12:06you don't need to. Correct.
- 12:08That 20 to 80 guideline is designed to minimize the voltage
- 12:11induced chemical strain on the cell structure, particularly on
- 12:15the cathode material at high voltages and the anode at very
- 12:18low voltages. It's important though to mention
- 12:21the exception here, LFP batteries.
- 12:23That's lithium iron phosphate if your vehicle uses LFP chemistry,
- 12:27which is becoming increasingly popular, especially in standard
- 12:30range models. Like some Teslas and Fords now.
- 12:32Exactly. LFP cells are inherently much
- 12:35more robust at high states of charge.
- 12:37They don't suffer the same level of stress near 100% as NMC cells
- 12:40do. So generally LFP batteries are
- 12:42much happier being charged to 100% regularly.
- 12:46That 2080 rule doesn't really apply in the same way.
- 12:48Good clarification. OK.
- 12:50So we have temperature, state of charge.
- 12:52What's the third big stress factor?
- 12:53The third one is the C rate, which basically refers to the
- 12:57current being drawn from Oregon pushed into the battery.
- 13:00High currents, whether they're being pulled out during really
- 13:03aggressive acceleration or pushed in during very fast
- 13:06charging, cause faster degradation.
- 13:09C rate. How is that defined exactly?
- 13:11It's defined simply as the operating current measured in
- 13:14amps, normalized by the nominal capacity of the cell measured in
- 13:18amp hours. So A1C rate means you're drawing
- 13:21enough current to theoretically discharge the entire battery in
- 13:24one hour. A 2C rate would discharge it in
- 13:26half an hour, and so on. Higher C rates mean higher
- 13:30currents. And why are high currents bad?
- 13:32Is it just heat again? Heat is a major part of it, yes,
- 13:36High currents inherently generate more internal heat
- 13:38within the cell due to resistance, what's called jewel
- 13:40heating. But it's not just heat.
- 13:43High currents can also increase mechanical strain inside the
- 13:45delicate structures of the electrode particles and
- 13:48potentially accelerate some of those unwanted side reactions
- 13:52like lithium plating on the anode, especially during
- 13:54charging at low temperatures. OK.
- 13:56And this is where our charging habits really come into sharp
- 13:58focus, right? The data shows that frequent use
- 14:01of direct current fast charging, or DCFC, those super fast
- 14:05chargers you find on highways. Right, the ones delivering 50
- 14:07kilowatt, 150 kilowatts, even 350 kilowatts now.
- 14:11That significantly impacts degradation rates compared to
- 14:13slower charging. Why is DCFC so much harder on
- 14:17the battery than say, my level 2 charger at home?
- 14:21Well, it's kind of a double way. I mean, especially if conditions
- 14:23aren't ideal. First, DCFC equipment uses
- 14:26extremely high currents by definition.
- 14:28That's how it charges fast. As we just discussed, high
- 14:31currents generate significant internal heat.
- 14:34Now, if you couple that high charging current with a hot
- 14:36ambient day, say you're fast charging in the summer heat
- 14:39after driving hard, the battery temperature can spike quite
- 14:42significantly. And we already know high
- 14:43temperature is enemy #1 so you're hitting it with high
- 14:45current and high heat simultaneously.
- 14:48Yeah. Level 2 AC charging, the kind
- 14:51you typically do overnight at home or at work, operates at
- 14:53much, much lower currents, maybe 610 kilowatts compared to 53150
- 14:59kilowatts. So it generates far less
- 15:01internal heat and is inherently much gentler on the battery
- 15:04chemistry. So it really is that classic
- 15:06tradeoff, the convenience of speed right now versus
- 15:10maximizing the batteries longevity over the long haul.
- 15:13Precisely. Using DCFC occasionally when
- 15:16needed for long trips is fine, that's what it's there for.
- 15:19But relying on it for daily charging, especially in hot
- 15:22climates, is generally advised against if you want to maximize
- 15:26lifespan. OK, that makes sense now.
- 15:27Here's a finding from the Geotab study that genuinely surprised
- 15:30me because it seems a bit counterintuitive at first
- 15:33glance. The data showed that vehicles
- 15:35with really high usage, driving lots of miles per year, do not
- 15:38necessarily experience significantly greater battery
- 15:40degradation than lower use vehicles.
- 15:43The difference in capacity loss they found was minimal,
- 15:45something like only .25% / 4 years.
- 15:47Yeah, that often surprises people.
- 15:49Why isn't simply using the battery more, putting more
- 15:51cycles on it, accelerating its decay more noticeably?
- 15:55This is a really critical distinction that helps clear up
- 15:57a major misconception. The mileage itself, or even just
- 16:01the sheer number of charge discharge cycles in isolation,
- 16:04isn't the sole primary driver of accelerated aging in many real
- 16:09world scenarios. The study concluded that the
- 16:11slight degradation acceleration they did observe in those high
- 16:14use vehicles was actually attributable to something else
- 16:19the need for more frequent DC fast charging to support that
- 16:22high mileage demand, especially if those high use vehicles were
- 16:25operating in hotter climates where the fast charging is even
- 16:28more stressful. So it's not the driving, it's
- 16:30the charging needed to enable the driving.
- 16:32Exactly. If a high use driver, maybe a
- 16:34taxi or a delivery van can consistently recharge using
- 16:38gentle overnight level 2 AC charging, the rate of battery
- 16:42degradation actually approaches that of a low use vehicle that's
- 16:46charged the same way. So the key take away here is
- 16:49actually quite positive for using EVs.
- 16:51High vehicle utilization is great.
- 16:53It's encouraged even to maximize the economic benefits and reduce
- 16:56the total cost of ownership. You just need to optimize the
- 16:59charging methods to avoid that daily reliance on high stress
- 17:02fast chargers whenever possible. That's a powerful clarification.
- 17:06So the issue isn't driving 50,000 miles a year.
- 17:09The potential issue is that driving 50,000 miles a year
- 17:11might force you to hit a fast charger four times a week, and
- 17:14that's what adds the extra stress, especially if it's hot
- 17:17out. You've nailed it.
- 17:18That's the new one. OK, one more factor from this
- 17:20section, Buffer effects. What are these?
- 17:22Right buffer effects. This is a really important
- 17:25engineering factor that manufacturers use to help manage
- 17:28all three of those stress factors, temperature, SoC, and
- 17:31current. Behind the scenes buffers are
- 17:34essentially the manufacturer's vital internal protection system
- 17:37for the battery. Automakers deliberately
- 17:39implement charge buffers to control the usable state of
- 17:42charge window that you, the driver, see and interact with.
- 17:45They effectively mask the battery's true chemical
- 17:48extremes. So when my dashboard says 100%,
- 17:51it's not really 100% of the battery's chemical capacity, and
- 17:55when it says 0%, it's not truly empty.
- 17:57Correct. There's usually a hidden buffer
- 17:59at the top, meaning 100% displayed might be only say 95%
- 18:02of the true chemical maximum, and there's a buffer at the
- 18:04bottom. So 0% displayed might still
- 18:06leave 5% or 10% of the actual chemical capacity remaining.
- 18:10These buffers prevent the driver from accidentally pushing the
- 18:13cells into those really high stress voltage zones, the
- 18:16absolute chemical 0% or 100%, which we identified earlier as
- 18:19being particularly damaging. OK, so a larger buffer means the
- 18:22manufacturer is protecting the battery more aggressively, and
- 18:25it also means less usable range for the driver from the same
- 18:28size pack. Exactly, it's another trade off
- 18:31made by the engineers this time. Less usable range day-to-day,
- 18:35but potentially slower degradation over the long term.
- 18:38Can you give an example of a car known for large buffers?
- 18:41The sources often highlight the original Chevrolet Volt plug in
- 18:45hybrid as a classic case study. It was well known in engineering
- 18:49circles for having very large top and bottom buffers on its
- 18:52battery pack compared to many other EVs of its time.
- 18:55And the result data showed the Volts battery exhibited slower
- 18:59than average degradation over its lifespan.
- 19:01It clearly demonstrated the long term benefit of sacrificing a
- 19:05bit of usable range upfront for that maximum protection,
- 19:08especially from those high SoC levels.
- 19:11Interesting. And this leads to a practical
- 19:13actionable tip for drivers, right?
- 19:15We can sort of replicate this buffer effect ourselves.
- 19:17Yes, absolutely. Most modern EVs allow you, the
- 19:21owner, to set a custom charge limit through the infotainment
- 19:24system or smartphone app. You can choose to charge to say
- 19:2875% or 85% for your daily needs instead of letting it go all the
- 19:33way to the displayed 100% every single time.
- 19:36So by setting that lower limit for daily charging.
- 19:38You are proactively keeping the battery out of that higher
- 19:41stress high voltage range near the top, even beyond the
- 19:44manufacturer's built in buffer. This can further extend the
- 19:48battery's chemical life over many years.
- 19:49And you just reserve that full 100% charge whatever the car
- 19:52allows. Only for those specific long
- 19:55road trips where you genuinely you need every single mile or
- 19:57kilometer range. Precisely charge it fully right
- 19:59before you depart for the long trip.
- 20:01But for daily use, stopping at 80% or 85% is often a great
- 20:05strategy for longevity, especially with NMC and CM
- 20:08chemistries. OK, that covers the big factors
- 20:10we see in the real world. Temperature, SoC management,
- 20:14charging speed, and buffers. Now let's zoom in.
- 20:17Let's go microscopic To really understand why these factors
- 20:21matter. Engineers use these incredibly
- 20:23detailed scientific models like the PDD models you mentioned,
- 20:26often combined with post mortem analysis, basically taking apart
- 20:30old batteries to see what failed.
- 20:32Exactly. These advanced physical models
- 20:34are essentially sophisticated computer simulations of
- 20:37everything happening inside the battery at a very fine level.
- 20:41They model how the lithium ions move back and forth, how the
- 20:44electrode particles might crack or change structure, how
- 20:47unwanted chemical layers form on surfaces.
- 20:50They allow researchers and engineers to pinpoint precisely
- 20:53which parts of the battery are degrading under different
- 20:55conditions, and crucially, why. We see that non linear
- 20:58degradation curve we talked about earlier, that initial
- 21:00steeper drop followed by slower aging.
- 21:03OK, it sounds incredibly complex.
- 21:05Let's try to simplify it. What are the main sort of
- 21:08chemical bad guys, the primary degradation phenomena identified
- 21:12by these models? The research generally points to
- 21:14three key degradation phenomena. The 1st and often considered the
- 21:19most dominant degradation mechanism, especially for
- 21:21calendar aging. That time based aging when the
- 21:24car is parked is something called solid electrolyte
- 21:27interface growth or SEI growth. SCI growth for the listener who
- 21:31isn't a battery chemist. What exactly is the SEI layer?
- 21:34Can we use an analogy? Yeah, let's try think of the SEI
- 21:37layer like maybe like a very thin layer of plaque or even
- 21:41scar tissue that gradually forms on the surface of one of the
- 21:44batteries key internal components, the negative
- 21:46electrode or the anode, usually made of graphite.
- 21:48OK, Plaque or scar tissue, How does it form?
- 21:50It forms because the liquid electrolyte, the substance that
- 21:53carries the lithium ions back and forth between the
- 21:55electrodes, isn't perfectly stable.
- 21:58Over time, especially at higher temperatures and higher states
- 22:01of charge voltages, the electrolyte slowly decomposes
- 22:04where it touches the anode surface.
- 22:06This decomposition process creates this thin, solid but
- 22:10generally electronically insulating layer.
- 22:12The SEI post mortem analysis where they cut open aged cells
- 22:16have confirmed this. Using techniques like Sims.
- 22:19They can see this layer gets thicker and covers the anode
- 22:22surface more uniformly as the battery ages.
- 22:25They even see excess oxygen trapped in it, confirming it
- 22:27comes from electrolyte breakdown.
- 22:29OK, so this layer forms and grows over time.
- 22:32Why is that bad? What are the consequences?
- 22:35It has two major negative consequences. 1st and most
- 22:38importantly for capacity loss, its formation leads directly to
- 22:41loss of lithium inventory or LL. Loss of lithium inventory LOI.
- 22:46Right, because lithium ions from the electrolyte and the
- 22:49electrode are actually consumed and permanently trapped in the
- 22:52process of forming that stable SEI layer.
- 22:55They become part of the layer itself and are no longer
- 22:58available to shuttle back and forth to store energy.
- 23:01This directly reduces the total amount of cyclable lithium in
- 23:05the battery, which is the primary cause of capacity fade.
- 23:08You're losing your active ingredient.
- 23:11Less usable lithium means less range.
- 23:13What's the second consequence? The second consequence is that
- 23:16the SEI layer itself, being solid and often electronically
- 23:19resistive, acts as a barrier. It slightly impedes the movement
- 23:24of lithium ions trying to get into and out of the anode.
- 23:27This increases the battery's internal resistance and
- 23:30decreases the conductivity of the electrolyte near the
- 23:32electrode surface. An increased resistance effects
- 23:35power output. Exactly.
- 23:37It directly impacts the battery's ability to deliver
- 23:40high power quickly, which can limit peak acceleration,
- 23:43especially as the layer thickens over many years or if it grows
- 23:46particularly fast due to high temperatures.
- 23:48So SEI growth is kind of a double whammy.
- 23:51It steals your lithium reducing range and it slows down ion
- 23:54movement reducing power. Makes sense why it's considered
- 23:57dominant. What's the second key mechanism?
- 24:00The second mechanism is really just LLI loss of lithium
- 24:03inventory. Considered more broadly, while
- 24:05SEI growth is a major cause of LLI, LLI itself is the
- 24:10overarching phenomenon. It accounts for all capacity
- 24:13loss associated with lithium becoming unavailable for
- 24:15cycling. This could be due to SEI
- 24:17formation as we discussed, but also potentially due to other
- 24:21side reactions happening elsewhere in the cell, or
- 24:23lithium getting trapped within the electrode structures or
- 24:26plating onto the anode surface under certain conditions like
- 24:28fast charging in the cold. LLI is essentially the primary
- 24:32metric for how much capacity the battery has lost.
- 24:35Think of LLI is losing the fuel itself.
- 24:37OK. LLI, is the outcome lost
- 24:39lithium, right? And the third mechanism.
- 24:41The third critical mechanism is loss of active material, or
- 24:44lamb. Now, if LLI is about losing the
- 24:46fuel, lithium, Lamb is about the physical degradation or chemical
- 24:50dissolution of the containers or structures that hold that fuel,
- 24:53the active materials in the electrodes themselves.
- 24:55So the electrodes themselves are breaking down.
- 24:57Essentially, yes, or becoming less effective.
- 25:00And interestingly, Lamb behaves differently depending on whether
- 25:04it's happening on the negative electrode anode or the positive
- 25:07electrode cathode. Let's start with Textilium
- 25:10dollar. That's loss of active material
- 25:12on the negative electrode. The PTD modeling and postmortem
- 25:15analysis often show this has a characteristic pattern, a
- 25:18relatively steep initial growth phase happening early in the
- 25:21battery's life, which is then often followed by a plateau
- 25:24where the rate of Textilium slows down significantly.
- 25:27A steep initial phase? Hang on, that sounds familiar.
- 25:30Is that connected to the initial steeper capacity drop we see in
- 25:33the real world data? Bingo, that's the current
- 25:36interpretation. This initial text delay is
- 25:38thought to be related to mechanical stress and micro
- 25:41cracking of the anode particles like graphite as they expand and
- 25:44contract during the very first charge discharge cycles.
- 25:48There might also be some initial restructuring or passivation
- 25:51happening as part of that settling in or stabilization
- 25:53process we mentioned earlier. The anode takes its biggest
- 25:56structural hit early on and then kind of stabilizes.
- 25:59Fascinating. So that initial drop isn't just
- 26:02SEI forming, it's also the anode structure itself adjusting,
- 26:06maybe even slightly damaging itself initially.
- 26:09That seems to be a key part of it, yes.
- 26:11The battery isn't just chemically settling in, it's
- 26:13also physically adjusting. OK.
- 26:15And what about the other side, Lamb on the positive electrode?
- 26:19Right. Text land, loss of active
- 26:20material on the positive electrode, the cacode.
- 26:24This tends to behave differently.
- 26:25The models and data suggest text land often grows more linearly
- 26:29over time rather than having that steep initial phase A.
- 26:33Steady continuous degradation. Yes, a steadier process and text
- 26:37land is known to be strongly promoted by high temperatures.
- 26:40The mechanisms here can include things like the slow dissolution
- 26:44of transition metals like nickel or cobalt from the cathode
- 26:47structure into the electrolyte, or perhaps structural changes
- 26:50like compa education mixing within the cathode crystal
- 26:53lattice. High heat definitely accelerates
- 26:56these positive electrodegradation pathways.
- 26:58OK. So summarizing the microphysics,
- 27:00SEI growth is key for calendar aging and causes.
- 27:04LLILLI is the overall loss of lithium.
- 27:07Lam is the physical breakdown, with the negative side lane
- 27:10having a big hit early and the positive side lamp degrading
- 27:14more steadily, especially when hot.
- 27:16That's a great summer, you've got the main players.
- 27:18Now let's connect this microscopic understanding back
- 27:20to the big environmental factors we discussed earlier,
- 27:23temperature and state of charge. Based on what the PDD model
- 27:26showed when simulating storage conditions, how does temperature
- 27:30affect these specific mechanisms?
- 27:32The models provide really clear confirmation here.
- 27:34Higher temperature acts as a systemic aggressor, accelerating
- 27:38pretty much all the bad stuff. Simulating storage at say 60° C
- 27:43versus a lower 45° C showed significantly accelerated LLI,
- 27:47meaning faster SEI growth and lithium consumption.
- 27:50It also accelerated that linear textile growth, more positive
- 27:53electrode breakdown and as expected, it led to faster
- 27:56formation of that SEI layer, resulting in a quicker decrease
- 28:00in the electrolytes conductivity near the anode.
- 28:02So high heat just makes everything worse faster attacks
- 28:05the battery on basically every chemical front.
- 28:07Exactly. It completely validates why
- 28:09effective thermal management like liquid cooling is so
- 28:13incredibly crucial for battery longevity in the real world.
- 28:16It's directly fighting these underlying degradation
- 28:19mechanisms. OK, temperature is bad.
- 28:21What did the models reveal about state of charge SoC during
- 28:25storage that might challenge some common assumptions?
- 28:28Well, the models definitely confirm that the extremes are
- 28:31detrimental, just as the empirical data suggests.
- 28:34Storing a battery at a displayed 100% SoC, which corresponds to a
- 28:38high cell voltage, leads to more severe self discharge and
- 28:42significantly accelerated degradation, particularly that
- 28:44SEI growth leading to LLI high voltage stresses the electrolyte
- 28:48and cathode. Conversely, storing it at a very
- 28:51low SoC, like 10% low voltage, resulted in very little
- 28:55degradation in the simulations. Chemically it's very stable
- 28:58state, although obviously not practical for daily use as you
- 29:00have no range. Right, so avoid 100% for
- 29:03storage. 10% is safe but useless.
- 29:05What about the middle ground, that 2080% roll?
- 29:08This is where the modeling provided a really interesting
- 29:10nuance, maybe even a slightly counterintuitive finding
- 29:13compared to the simple rule of thumb.
- 29:16The P2D model analysis found no significant difference in the
- 29:19key degradation indicators. Specifically looking at TE, XL,
- 29:23AM, P/E, A positive electrodegradation, and the
- 29:26decrease in electrolyte conductivity factor related to
- 29:29SCI impact between storing a battery at 50% SoC versus
- 29:33storing it at 80%. So C provided the temperature
- 29:36was kept constant. Wait a second, let me process
- 29:38that. No significant difference in
- 29:40degradation between 50% and 80% storage according to the model.
- 29:43At least for those specific degradation modes under instant
- 29:46temperature storage conditions, yes, the difference was
- 29:48negligible in their simulations. But we're constantly told,
- 29:52almost religiously, to stick to that 20% to 80% rule for daily
- 29:56charging. If the model shows the actual
- 29:58chemical strain difference between, say, 50% and 80%
- 30:01storage is minimal, why is that 80% limit pushed so hard?
- 30:04Is it just about avoiding 100%? That's an excellent question,
- 30:07and it really gets to the heart of translating complex
- 30:10engineering strategy into simple consumer advice.
- 30:13The model confirms the relationship isn't linear.
- 30:16The real accelerated damage, especially from a voltage stress
- 30:20perspective, occurs at the absolute extremes, getting close
- 30:24to that true chemical 100% high voltage and dropping below maybe
- 30:281020% very low voltage. The chemical stress difference
- 30:32between sitting at 50% SoC and sitting at 80% SoC is relatively
- 30:37small in comparison, particularly for calendar aging
- 30:40mechanisms when temperature is controlled.
- 30:41So why the 80% rule then? The 80% rule is often maintained
- 30:45and communicated as a simple, easy to remember and
- 30:47conservative target. Its main benefit is that it
- 30:50reliably keeps the average driver safely away from that
- 30:53much more critical 9100% zone where SEI growth really starts
- 30:57to ramp up significantly due to the higher voltage, especially
- 31:00during prolonged storage or calendar aging.
- 31:03Think of it more as a practical guideline for risk management
- 31:05rather than a hard chemical Cliff edge of exactly 80.0%.
- 31:09It builds in a safety margin. That makes a lot more sense.
- 31:12So the 80% limit feels less like a strict chemical law you must
- 31:16obey and more like a sensible guardrail to prevent you from
- 31:19getting too close to the real danger zone near 100%.
- 31:21Exactly. It also highlights that from a
- 31:24pure chemical degradation standpoint, if you only need,
- 31:27say, 50% charge for your daily commute, charging up to 70%
- 31:32instead of rigidly stopping at 80% probably gains you very
- 31:35little extra battery life protection, but stopping earlier
- 31:38does save you charging time and energy consumption.
- 31:41Right. The real protective effort needs
- 31:42to be focused on avoiding high temperatures and consistently
- 31:46parking the car overnight at 100%, rather than obsessing over
- 31:4975% versus 80%. That's a very practical way to
- 31:52look at it based on these deeper models.
- 31:54Focus on the big hitters, heat and the true top end of the
- 31:57charge. OK, this understanding of the
- 31:59mechanisms in the trade leads us perfectly into the final
- 32:02section, how engineers try to manage all this in the car
- 32:06itself. Yes, these insights feed
- 32:08directly into the engineering challenge, which is essentially
- 32:12a closed loop control problem managed by the battery
- 32:14management system, the BMS. You can think of the BMS as the
- 32:18tireless traffic cop living inside your battery pack.
- 32:22Engineers using the BMS are constantly trying to solve a
- 32:25tricky multi objective optimization problem.
- 32:28They have competing goals. They want to minimize battery
- 32:31aging. Let's call that cost function J
- 32:33Life dollars. But at the same time, they need
- 32:36to maximize everything the driver actually wants and
- 32:39expects from the car. Performance.
- 32:40Exactly. Maximize speed and acceleration,
- 32:43maximize the available driving range, and minimize the time it
- 32:46takes to charge. And as we've established
- 32:48throughout this conversation, those goals are fundamentally in
- 32:51conflict, aren't they? If you push for maximum
- 32:54performance, you inevitably use higher currents.
- 32:58Maybe push the state of charge to its limits?
- 32:59And that stresses the battery, accelerating that the aging
- 33:02process. You can't have it all all the
- 33:04time, so. How does the BMS actually try to
- 33:06balance this trade off in real time while you're driving or
- 33:09charging? What levers does it pull?
- 33:11The BMS primarily uses 2 main control variables, 2 main noms.
- 33:15It can turn to manage this dynamic balancing act.
- 33:19First, it can actively set the maximum current that's allowed
- 33:22to be drawn from the battery at any given moment, particularly
- 33:25during acceleration. And 2nd, it can actively limit
- 33:28the depth of discharge DoD by managing the charging process
- 33:32and the usable SoC window, essentially enforcing those
- 33:36buffers we talked about and potentially adjusting them.
- 33:38OK, IMAX and DoD limits. Let's look at the simulations
- 33:42discussed in the research. Specifically for fully electric
- 33:44vehicles Fev's. They modeled driving an FEV over
- 33:48a simulated 30,000 kilometer cycle under different control
- 33:51strategies. Well, first they tested the
- 33:53effect of limiting an IMAX, the maximum current.
- 33:56For example, they put a constraint on the system saying
- 33:58never draw more than .75 C What did that do?
- 34:02They found that limiting the current did provide a small
- 34:04benefit for battery life. It limited the overall capacity
- 34:07loss, saving about .25% of capacity over that 30,000
- 34:11kilometer simulation compared to letting the current run free, a
- 34:14measurable but small improvement.
- 34:15OK, 1/4 of a percent saved, but what was the cost?
- 34:18The cost was an immediate and noticeable degradation of the
- 34:21drivers living experience. Limiting IMAX directly
- 34:24translates to limiting maximum acceleration.
- 34:27The car feels more sluggish when you floor the pedal.
- 34:29It also inherently slows down the fast charging process as you
- 34:33can't push current in as quickly either.
- 34:35Right. Less fun to drive, takes longer
- 34:36to charge. And interestingly, the
- 34:38simulation also showed that if the driving cycle itself was
- 34:42naturally very smooth, maybe just gentle Hwy.
- 34:46Cruising, where high currents are rarely demanded anyway, then
- 34:49the benefit of actively limiting iMac became almost you sacrifice
- 34:53driving dynamics for a tiny battery life game, and only if
- 34:57you drive aggressively in the first place.
- 34:58OK, so limiting current has a small benefit, but a definite
- 35:01performance downside. What about the other lever
- 35:04limiting the depth of discharge? Right, limiting DoD.
- 35:07This means actively reducing the maximum usable capacity window,
- 35:11for instance, only allowing the battery to be used down to, say,
- 35:1450% of its total pack capacity between charges.
- 35:17In the FEV simulations, this strategy showed a much more
- 35:20powerful effect on preserving battery life.
- 35:23Limiting the DoD in this way saved approximately 1% of
- 35:26capacity over that same 30,000 kilometer simulation horizon. 1%
- 35:30save by limiting the usable charge window, compared to only
- 35:34.25% save by limiting the peak current.
- 35:37That's four times the impact. It is significantly more
- 35:40effective in these simulations and this highlights a really
- 35:43crucial engineering point, especially when comparing fully
- 35:45electric vehicles, Fev's to older hybrid vehicles.
- 35:49HEVSFEV's, particularly modern ones, typically have very large
- 35:53battery packs relative to the power demands of the motor, much
- 35:56larger than the small packs and hybrids.
- 35:58Because the overall pack is so large, the individual cells
- 36:01within that pack actually experienced relatively smaller C
- 36:03rates currents even during quite high acceleration events.
- 36:06The load is spread across more cells.
- 36:08Exactly. Therefore, in these large pack
- 36:11Fevs, actively limiting the DoD, managing that charge window
- 36:15carefully keeping it away from the extremes tends to have a
- 36:19significantly more evident and powerful effect on slowing down
- 36:22battery aging then trying to clamp down hard on the maximum
- 36:26current draw, which might already be relatively modest at
- 36:29the cell level. That makes sense.
- 36:31So for the owner of a typical modern EV with a decent sized
- 36:35battery pack, outside of managing temperature,
- 36:37controlling your daily charging window like that 80% limit we
- 36:40discussed is probably the single biggest lever you have direct
- 36:43influence over for longevity. Based on this research,
- 36:46absolutely temperature and SoC window management seem to be the
- 36:49dominant factors accessible to the driver for Fe VS.
- 36:52OK, now we get to to what sounds like the most sophisticated
- 36:55part, the optimal control strategy that the researchers
- 36:58derived. They weren't just testing fixed
- 37:00limits, they tried to find the absolute best way to manage the
- 37:03battery over its whole life. That's right, they didn't just
- 37:06look for a single set it and forget it limit for IMAX or DoD.
- 37:10Instead they created a theoretical optimal benchmark.
- 37:13They did this by mathematically weighting the performance
- 37:16indices, speed range, charge time against the battery life
- 37:21over a very long simulated horizon 200,000 kilometers.
- 37:25So trying to find the perfect balance over the long haul, what
- 37:28did this optimal strategy look like?
- 37:30Was it just tighter limits all the time?
- 37:31No, and this is the really key result.
- 37:34The optimal strategy turned out not to be a constant set of
- 37:37limits. It was a dynamic, time varying
- 37:39control sequence. The required level of protection
- 37:42changes over the vehicle's life. And here is the crucial insight.
- 37:45The optimal strategy dictates that battery protection needs to
- 37:48be heavily front loaded. Front loaded, meaning you
- 37:51protect it more when it's new. Precisely the optimal control
- 37:54requires the BMS to exert a greater control effort, be more
- 37:57restrictive in the first part of the vehicles life.
- 38:00Specifically, the simulation suggested the 1st 50,000
- 38:03kilometers around 30,000 miles are the most critical.
- 38:07OK. What is that greater control
- 38:09effort look like during those first 50,000 kilometers?
- 38:12During this initial sensitive period, the optimal strategy
- 38:16derived in the paper aggressively limits the usable
- 38:18depth of discharge DoD, often keeping it below 50% of the
- 38:22packs opacity, and it also limits the maximum allowed
- 38:25current, quite strictly keeping it below a moderate 2 point OC.
- 38:29Wow, DoD below 50% in IMAX? Below 2C?
- 38:33That sounds quite restrictive compared to normal driving.
- 38:35Why is such aggressive protection needed right at the
- 38:37beginning? The rationale ties directly back
- 38:39to the microphysics we spent time on in Section 3.
- 38:43Remember how we discussed that the agent degradation rate is
- 38:45significantly higher during the first thousands of ampere hours
- 38:49processed by the battery? Yeah, due to that initial steep
- 38:51SEI formation and the Lamb stabilization, the anode
- 38:54cracking. Exactly because those initial
- 38:57irreversible degradation processes are happening most
- 39:00rapidly right at the start, the optimal control strategy
- 39:03recognizes this. By protecting the battery really
- 39:06hard and early during that highly sensitive initial
- 39:09stabilization period, you prevent the excessive
- 39:12irreversible consumption of lithium LLI and minimize that
- 39:16initial structural damage AMM. The benefits of this early
- 39:21protection then accumulate and pay dividends over the entire
- 39:24remaining life of the vehicle. You're preserving more capacity
- 39:28and health from the outset. OK.
- 39:30That's a powerful concept. Protecting the car when it's
- 39:33brand new is far more valuable in terms of long term capacity
- 39:37than applying the same level of protection when it's, say, 5
- 39:39years old and already stabilized.
- 39:41That's the core finding of the optimal control analysis.
- 39:44And what happens after that initial aggressive period after
- 39:46the 1st 50,000 kilometers in the simulation?
- 39:48Does the BMS just keep those tight limits forever?
- 39:51No, the control effort actually relaxes quite significantly
- 39:54after that initial 50,000 kilometers mark.
- 39:57The underlying degradation curve has typically flattened out.
- 40:00The battery is stabilized, so the optimal control sequence
- 40:03then allows the DoD limit to relax, settling back somewhere
- 40:07around 50% to 60% for the remainder of the vehicle's life.
- 40:11And the maximum current limit also relaxes, allowing peaks up
- 40:14to around 2.5 C. The simulation confirmed that
- 40:18this dynamic strategy, tight control, early relaxed control,
- 40:21later performed demonstrably better in preserving capacity
- 40:24over the entire 200,000 kilometer horizon compared to
- 40:28just applying fixed moderately relaxed constraints over the
- 40:31whole life. So it's not just about setting
- 40:32limits, it's about setting the right limits at the right time.
- 40:34Exactly, it's a crucial finding for engineers designing these
- 40:37BMS strategies. You potentially save the most
- 40:40long term capacity by front loading the protective effort.
- 40:43This has been an incredibly detailed journey.
- 40:45We've gone from the macro level data of 10s of thousands of cars
- 40:48right down to the microphysics of particle cracking and SEI
- 40:51layers and back up to sophisticated engineering
- 40:54control strategies. I think the key take away for
- 40:57you, the listener is really clear now your EV battery's life
- 41:02isn't just some lottery. It's not just luck of the draw.
- 41:05It's a direct function of informed choice you make,
- 41:08especially regarding temperature exposure and charging habits,
- 41:10combined with the highly advanced, often invisible
- 41:13control strategy running inside the car's BMS and maybe the
- 41:17shortcut to being well informed. The core principles to
- 41:19internalize are these three. First, prioritize vehicles with
- 41:24good thermal management, ideally liquid cooling, especially if
- 41:27you're in a hot climate. Second, be mindful of charging.
- 41:30Limit routine reliance on DC fast charging to avoid
- 41:33unnecessary heat and current stress.
- 41:35Use Jetler Level 2 charging when possible.
- 41:38And 3rd, realize that the early life of the battery, maybe those
- 41:41first 50,000 kilometers or 30,000 miles, is the most
- 41:44sensitive period where careful SoC management potentially
- 41:47yields the biggest long term benefits.
- 41:49Absolutely. And building directly on that
- 41:51last point, that finding about the dynamic optimal control
- 41:55strategy, the fact that the degradation rate is highest in
- 41:57the first few years and the optimal way to maximize lifespan
- 42:01involves potentially limiting performance quite aggressively
- 42:04early on. Well, this raises a really
- 42:07important and quite provocative question that car manufacturers
- 42:11are undoubtedly grappling with right now, which is does the
- 42:13consumer need a new kind of interface?
- 42:16A new choice? Since optimal protection,
- 42:18especially early on, requires limiting performance like
- 42:21acceleration and usable range, lower DoD limits, should you,
- 42:24the driver, actually be given the ability to choose your
- 42:27personal trade off? Mean let the driver decide.
- 42:30Yeah, should you be able to say, you know what, for the 1st
- 42:3250,000 kilometers, I prioritize maximum performance and
- 42:36accessing the full advertised range.
- 42:38I accept the consequence that this might accelerate the
- 42:41battery's aging. Or conversely, should you be
- 42:44able to choose? No, I want to optimize for
- 42:47maximum possible lifespan right from day one.
- 42:50I'm willing to accept slightly reduced acceleration and maybe a
- 42:53software limited daily range for the first couple of years to
- 42:56achieve that. Performance versus persistence
- 42:58chosen by the customer. Exactly.
- 43:00That choice, that fundamental conflict between getting the
- 43:03most out of the card now versus making it last as long as
- 43:06possible is the core engineering trade off.
- 43:09And the question is becoming how much of that trade off decision
- 43:13should be passed along from the engineers to you, the customer?
- 43:16That is a fascinating and definitely provocative thought
- 43:19to ponder next time you look at that state of charge meter, a
- 43:22potential future direction for EV interfaces.
- 43:25Well, thank you for joining us today for this really deep dive
- 43:28into the science and management of EV batteries.
- 43:30We hope you found it useful. If you appreciate having these
- 43:33complex topics broken down into actionable, clear knowledge,
- 43:37please do consider following the channel for more explorations
- 43:39like this one. We look forward to diving into
- 43:41another topic with you again soon.