Daniel sent us this one and I think it hits on something a lot of homeowners feel but can't quite articulate. You drop serious money on a home battery, the spec sheet says thirteen point five kilowatt-hours, and by year three it feels like you're getting maybe seventy percent of what you paid for. He's asking why this gap exists, whether it's actually degradation or something else, and what the manufacturers aren't telling you upfront. Honestly, there's more to this than just chemistry.
It's one of those topics where the online discussion is split between people who are angry and people who are reciting spec sheets, and neither side is actually explaining what's happening in the middle. By the way, today's episode is powered by DeepSeek V four Pro. Which I mention now because we're about to dive into battery electrochemistry and I will absolutely forget later.
So where do we even start with this? Because I think most people assume the battery's just losing capacity, like a phone battery after a few years, and that's not the full picture.
Right, and the phone battery comparison is actually the first thing we need to knock down because it's misleading in a very specific way. When your phone battery degrades, you feel it directly. The phone dies faster. With home batteries, there's a whole layer of software between you and the actual cells, and that software is making decisions that can look a lot like degradation even when the cells are fine.
You're saying the battery management system is part of what people perceive as capacity loss.
And not just the BMS. The inverter, the round-trip efficiency losses, the way the manufacturer defines usable capacity versus total capacity, the temperature where the battery lives. All of these compound. A lithium iron phosphate cell might lose two to three percent of its actual chemical capacity per year under ideal conditions. That's the real degradation number. But what a homeowner sees at the wall outlet can look like fifteen or twenty percent less by year three, and they blame the cells when half of that is system-level losses that were there on day one.
There's a difference between nameplate capacity and what you can actually pull out, and that difference gets blamed on degradation.
And here's the thing. Most manufacturers advertise the total capacity, but the usable capacity is what you actually get. The Tesla Powerwall three, for example, has thirteen point five kilowatt-hours of total storage, and they've actually closed that gap in recent generations. But earlier models, and many competing systems, have a buffer. The LG Chem RESU ten H has a total capacity of nine point eight kilowatt-hours but a usable of nine point three. Enphase's IQ Battery five P advertises five kilowatt-hours but the usable is four point nine six. These are small buffers, but they're real.
Those buffers exist to protect the cells, right? So it's not a scam, it's protective.
It's protective, but it also means the number on the box isn't the number you get to use. And the buffer often grows over time as the battery ages, because the BMS becomes more conservative. So you might start with ninety-five percent usable, and by year three the BMS has decided it needs to hold back an extra five percent to prevent deep discharge damage. The cells haven't lost that capacity, the system has just locked it away.
That's actually fascinating. The battery's not losing capacity, the software is walling it off for safety, and the homeowner just sees less run time.
And the warranty language is where this gets really interesting. I was looking at a breakdown on Clean Energy Reviews about how these warranties are structured, and almost none of them guarantee usable capacity at the wall. They guarantee the cells will retain a certain percentage of original capacity under specified test conditions, which are not your garage in Phoenix in August.
Okay, let's talk about those warranty numbers. What's typical?
The industry standard, across Tesla, LG, Enphase, Sonnen, most of the major players, is a warranty that guarantees seventy percent of original capacity after ten years or a certain number of cycles, whichever comes first. Tesla's is ten years and unlimited cycles. Enphase is ten years or four thousand cycles. LG is ten years or a specific energy throughput number that varies by model. And here's the key detail. That seventy percent guarantee is measured at the cell level under controlled lab conditions. It's not measured at your inverter output on a cold January morning.
Which means your real-world number could be worse and you have no warranty claim.
And the degradation curve isn't linear. Most lithium batteries lose more capacity in the first year or two than they do in years three through seven. So a battery might drop to ninety percent in year one, eighty-five percent by year three, and then the curve flattens. That early drop feels alarming, but it's actually expected. The solid electrolyte interphase layer, the SEI layer, forms and stabilizes during those early cycles, and that formation consumes a small amount of lithium.
SEI layer formation. That's the actual chemical process.
That's the actual chemical process. When you first cycle a lithium battery, the electrolyte reacts with the anode and forms this thin passivation layer. It's essential for long-term stability, but it consumes some active lithium in the process. That's why you see that initial drop. It's not a defect, it's a necessary part of the battery's life cycle. But manufacturers don't exactly put that in the marketing brochure.
We've got SEI formation eating some capacity early, we've got the BMS becoming more conservative over time, and we've got the gap between nameplate and usable capacity that was there from the start. What else is contributing to that perceived underperformance by year three?
And this is one where installation location matters enormously. Lithium iron phosphate batteries, which is what most home systems use now, have a sweet spot between about fifteen and thirty-five degrees Celsius. Outside that range, the internal resistance increases and you lose usable capacity. If your battery is in an uninsulated garage in a climate that hits freezing or above forty Celsius regularly, you're losing capacity seasonally that has nothing to do with degradation. It's temporary, but it's real.
Most people don't have climate-controlled battery rooms. They have a garage wall or an exterior mount.
And the battery's own thermal management system consumes power. The Powerwall has a liquid thermal management system. It will heat or cool itself to stay in the safe operating range, and that power comes from the battery itself or from the grid. That's parasitic load, and it's another small percentage that the homeowner never sees itemized. They just see less power available.
The battery's using its own power to keep itself comfortable.
Which is smart engineering. You want the battery to protect itself. But it adds to the gap between what you think you bought and what you actually get.
Let me ask about something I've seen people complain about online. People say their battery capacity seems fine but the inverter can't deliver the power they need, so they can't actually use the stored energy when they want to.
This is a huge one, and it's probably the most misunderstood spec in home energy storage. Capacity is measured in kilowatt-hours. That's how much energy is stored. But power output is measured in kilowatts. That's how fast you can pull that energy out. A battery might have thirteen point five kilowatt-hours stored, but if the inverter is rated at five kilowatts continuous, you can't run a seven-kilowatt load even if the battery is completely full. You'll hit the inverter limit.
People interpret that as the battery being depleted or degraded when it's actually just the power ceiling.
And inverter limits get more restrictive over time in some systems because the BMS will throttle power output as the battery ages or as temperature rises. So a battery that could deliver seven kilowatts peak when new might be capped at five kilowatts by year three because the BMS has decided the cells need more conservative treatment. The energy is still there, you just can't access it as quickly.
That's almost more frustrating than actual capacity loss. The energy exists, it's just behind a locked door.
It's one of the reasons why I tell people, when you're shopping for a battery, look at the continuous power rating and the peak power rating, not just the kilowatt-hours. The Powerwall three has a continuous output of eleven point five kilowatts, which is enough to run most home loads including an air conditioner. But many competing systems are in the five to seven kilowatt range, which means you're going to have to manage your loads carefully during an outage. And that management can feel like the battery isn't doing its job.
We've covered SEI formation, BMS conservatism, nameplate versus usable, temperature effects, parasitic loads, and inverter limits. Is there anything else in the system that's eating into that perceived capacity?
Round-trip efficiency. Every time you charge and discharge a battery, you lose energy. For lithium-ion home batteries, the round-trip efficiency is typically around ninety to ninety-five percent. The Powerwall three claims ninety percent. So if you put ten kilowatt-hours in, you get nine out. That loss is consistent and it's not degradation, but it reduces the effective capacity of the system day one. And if the battery is cycling daily, that ten percent loss compounds in terms of what the homeowner perceives as available energy over time.
That's just physics, right? You can't get around it.
You can't. It's heat loss in the power electronics and the cells. Some systems are better than others. DC-coupled systems, where the solar panels connect directly to the battery without an extra AC-to-DC conversion step, tend to have higher round-trip efficiency. AC-coupled systems have more conversion losses. The difference can be three to five percentage points, which adds up over thousands of cycles.
The architecture of the system matters. And most people don't even know whether their system is AC or DC coupled.
Most installers don't volunteer that information. They sell you a package and you trust that it's optimized. But an AC-coupled system with a ninety percent efficient inverter and a ninety percent efficient battery charger is losing energy at two separate conversion points. Your effective round-trip could be eighty-one percent, and you'd never know unless you measured it at the meter.
Okay, let's talk about something that's been in the news. There's a lot of discussion about second-life EV batteries being repurposed for home storage. Does that change the degradation picture?
It does, and not necessarily in a good way. Second-life batteries have already gone through hundreds or thousands of cycles in a vehicle, where they've been subjected to high charge and discharge rates, temperature extremes, and vibration. By the time they're repurposed for home storage, they might already be at eighty percent of original capacity or lower. And the degradation curve for an aged battery is different from a new one. You might see a steeper decline in those first few years of home use because the SEI layer is already thick and the cell is already stressed.
The bargain price on a second-life system might not be such a bargain if you're losing capacity faster.
It depends on the application and the warranty. Some companies are very upfront about what you're getting. But the warranty on a second-life system is typically shorter and the capacity guarantee is lower. You might get a five-year warranty with a sixty percent capacity guarantee instead of a ten-year warranty with seventy percent. You need to do the math on cost per usable kilowatt-hour over the warranty period, not just the upfront price.
That's a good framework. Cost per usable kilowatt-hour over time.
That's the metric that almost nobody calculates when they're buying. They look at the total kilowatt-hour number on the spec sheet, multiply by their electricity rate, and think they've got a payback period. But if your usable capacity drops twelve percent in the first three years, and your round-trip efficiency is eighty-seven percent, and your inverter is capped lower than you thought, your actual payback is longer.
Let's get into some specific numbers. You mentioned the Clean Energy Reviews breakdown. What's the actual degradation data look like for the major brands?
There isn't a lot of independent long-term data publicly available, which is part of the problem. The manufacturers publish their own test data, and it shows degradation of roughly one to two percent per year for LFP chemistry under ideal conditions. But real-world data from early adopters, and this is anecdotal gathered from forums and some smaller studies, suggests that actual usable capacity decline is closer to two to four percent per year in the first few years, then levels off. The difference between lab and real world is temperature, cycling patterns, and depth of discharge.
Depth of discharge is another thing people don't think about. If you're cycling the battery from a hundred percent down to ten percent every day, that's harder on the cells than cycling between eighty and thirty percent.
Most LFP batteries are rated for around four thousand to six thousand cycles at eighty percent depth of discharge. If you go to a hundred percent depth of discharge regularly, that cycle life can drop by twenty to thirty percent. The BMS tries to protect against this by building in that buffer, but if you're consistently running your battery to empty, you're accelerating degradation.
The user's behavior matters too. It's not just the hardware.
This is where the software side gets interesting. Some systems, like Tesla's, use predictive algorithms to manage charging and discharging based on your usage patterns and weather forecasts. They'll avoid charging to a hundred percent if they predict you won't need it, or they'll hold back some capacity for predicted peak demand. That's smart for longevity, but it can also make the battery appear to have less capacity than it does, because the software is making decisions the homeowner didn't explicitly authorize.
The battery's capacity is being managed by an algorithm that thinks it knows better than you do.
Sometimes it does. But the transparency isn't there. You open the app, it says seventy-two percent, you assume that's seventy-two percent of thirteen point five kilowatt-hours, but it might be seventy-two percent of a reduced usable window that the algorithm has already adjusted. You're not seeing the full picture.
What about calibration? I know with phones, sometimes the battery percentage gets out of sync with the actual state of charge, and a full discharge and recharge fixes it. Does that happen with home batteries?
It does, and it's one of the simplest explanations for perceived capacity loss. The BMS estimates state of charge based on voltage and coulomb counting, and over time those estimates can drift. If the battery hasn't been fully charged and fully discharged in a while, the BMS might think the capacity is lower than it actually is. Some manufacturers recommend a full calibration cycle every few months. But most homeowners never do it because they're never told to.
Some of these "my battery lost twenty percent" complaints might be solved by just running a calibration cycle.
Not all of it, but some portion. And that's frustrating because it's a software problem masquerading as a hardware problem.
Let's talk about the chemistry differences. LFP versus NMC. Does one degrade faster?
LFP, lithium iron phosphate, degrades slower than NMC, nickel manganese cobalt. That's one of the main reasons the home storage industry has shifted almost entirely to LFP. LFP has a longer cycle life, better thermal stability, and doesn't use cobalt, which is expensive and has supply chain issues. The trade-off is lower energy density, but for a stationary home battery, weight and volume don't matter as much as they do in a car. LFP can do four thousand to six thousand cycles to eighty percent capacity, versus NMC which might do two thousand to three thousand cycles to the same point.
If you bought an NMC system a few years ago, you're seeing more degradation than someone buying an LFP system today.
A lot of the early home battery systems were NMC. The original Powerwall one and two were NMC. The Powerwall three switched to LFP. The LG Chem RESU ten was NMC, the newer RESU prime line is LFP. So there's a generation of early adopters who have NMC chemistry and are seeing steeper degradation curves, and they're comparing their experience to the current marketing that's based on LFP longevity.
That's a mismatch that's going to cause a lot of frustration as those early systems age.
The warranty replacement terms don't always make you whole. Most warranties are prorated or they replace your battery with a refurbished unit that has similar remaining capacity, not a new one. So you might have a battery at seventy-two percent capacity after eight years, file a warranty claim because the guarantee is seventy percent, and they send you a refurbished unit at seventy-four percent. You haven't gained much.
That's brutal. You're limping along for years waiting for it to drop below the warranty threshold, and the fix barely improves things.
The warranty claim process itself can be a nightmare. You need to provide data logs proving the capacity is below the threshold, and the manufacturer may argue that your installation conditions, your cycling patterns, or your inverter are to blame. The burden of proof is on the homeowner, and most people don't have the technical knowledge or the monitoring equipment to make a compelling case.
We've got multiple layers here. Real chemical degradation, BMS buffers that grow over time, temperature derating, inverter caps, calibration drift, and warranty structures that make it hard to get relief. Is there anything the industry is doing to address the transparency problem?
There's a push for standardized testing and reporting, something like the EPA fuel economy sticker but for batteries. The California Energy Commission has been working on a battery performance labeling standard. And some third-party testing labs are starting to publish independent degradation data. But we're years away from having the kind of transparent, standardized metrics that would let a homeowner accurately predict what their battery will deliver in year three, year five, year ten.
In the meantime, what should someone actually do if they're shopping for a battery right now and they want to avoid this gap?
First, look at the warranty's fine print. Don't just look at the headline seventy percent, ten years. Look at what conditions void the warranty. Look at whether the capacity measurement is at the cell or at the inverter output. Look at the ambient temperature range. If you live in Phoenix and the warranty is void above forty Celsius ambient, you're going to have a problem.
The installation location matters enormously, like you said.
Put the battery in a conditioned space if you can. A basement, a utility room inside the thermal envelope of the house. Not a garage that hits a hundred and ten degrees in August. Not an exterior wall in direct sunlight. The installation cost might be higher, but the capacity you preserve over ten years will more than pay for it.
What about the inverter sizing? You mentioned looking at continuous power, not just capacity.
Map out what you actually want to run during an outage. Not everything in your house. The critical loads. Refrigerator, lights, internet, maybe a well pump or a medical device. Add up the wattage of those loads and make sure the battery's continuous power rating can handle the surge when they all kick on at once. A fridge compressor starting can pull three times its running wattage for a second. If your inverter can't handle that surge, the battery will shut down to protect itself, and you'll be sitting in the dark thinking the battery is dead when it's actually just protecting itself from an overload.
That's another perceived failure that's actually a protection mechanism.
It happens all the time. Someone installs a battery thinking they can run their whole house, the power goes out, they turn on the air conditioner, and the system trips. They blame the battery. The battery did exactly what it was designed to do.
We've been talking mostly about lithium-ion, but there are other chemistries entering the home market. Flow batteries, saltwater batteries. Do they have the same degradation profile?
Flow batteries are a completely different beast. They don't degrade in the same way because the energy is stored in liquid electrolytes in external tanks, not in solid electrodes. The electrolyte can be replaced or replenished, and the cycle life is essentially unlimited. But the energy density is much lower, the systems are physically larger, and the upfront cost is higher. They're more of a commercial and industrial product right now. For residential, lithium is going to dominate for the foreseeable future.
Aquion tried that. They went bankrupt in twenty-seventeen. The chemistry works, it's very safe and environmentally friendly, but the energy density is terrible and the power output is low. They couldn't compete with lithium on cost per usable kilowatt-hour, even accounting for lithium's degradation.
Lithium's the game for now, and the question is how to manage the degradation that comes with it.
I think the most important thing for homeowners to understand is that not all capacity loss is degradation. Some of it is temporary, some of it is software, some of it is environmental, and some of it is just the difference between the marketing number and the real number. If you understand all of those factors, you can make a much more informed decision about what to buy and how to install it.
The marketing number versus the real number. That's really the heart of Daniel's question. The spec sheet says one thing, the experience says another, and the gap is filled with chemistry and physics and software that nobody explains.
The gap is also filled with a lot of angry forum posts from people who feel like they got bait-and-switched. And I understand the frustration. You spend ten or fifteen thousand dollars on a system that's supposed to give you energy independence, and three years later you're getting less than you expected. But a lot of that gap was predictable if you knew what to look for.
Which brings us back to the transparency problem. If the information exists but it's buried in warranty fine print and technical spec sheets that most homeowners never see, is the industry actually being deceptive or just bad at communication?
I think it's a mix. Some manufacturers are genuinely trying to communicate honestly. Tesla's spec sheets are relatively clear about usable versus total capacity. Enphase publishes detailed degradation curves. But the solar installers who are actually selling these systems to homeowners often oversimplify. They quote the headline number, they quote a payback period based on that number, and they don't get into the nuances. The homeowner signs the contract thinking they're getting thirteen point five usable kilowatt-hours for ten years, and that's not what the warranty actually promises.
The installer is a big part of the problem.
The installer is the interface between the manufacturer and the customer, and a lot of them are sales organizations first and technical consultants second. They're incentivized to close the deal, not to explain SEI layer formation and round-trip efficiency losses.
Alright, let's put a number on it. If someone buys a typical LFP home battery today, thirteen point five kilowatt-hours nameplate, what should they actually expect to get out of it in year three, accounting for everything we've discussed?
Let's walk through it. Start with thirteen point five total. Usable is probably around thirteen point two after the buffer. Round-trip efficiency at ninety percent means you get about eleven point nine out for every thirteen point two you put in. That's not capacity loss, that's conversion loss, but it affects your effective runtime. Then you've got maybe three percent real chemical degradation per year for the first three years, so you're at about ninety-one percent of original capacity, which brings your usable down to about twelve. Then temperature derating, if you're in a hot climate, might knock off another three to five percent seasonally. Parasitic loads for thermal management, another one to two percent. And the inverter limit might cap your peak power but doesn't reduce total energy, so that's more of a runtime constraint than a capacity constraint.
All in, you're looking at maybe ten and a half to eleven kilowatt-hours of actual delivered energy from a thirteen point five nameplate system. That's about seventy-eight to eighty-one percent of the advertised number.
That's in year three under real-world conditions. And that number isn't necessarily a problem if you planned for it. If you sized your system assuming you'd get thirteen point five and you're getting eleven, you're going to be disappointed. If you sized it assuming eleven, you're fine. The problem is that almost nobody sizes it assuming eleven.
Because the installer told them thirteen point five.
Because the installer told them thirteen point five and the payback calculation used thirteen point five.
The actionable advice here is, when you're shopping, take the nameplate number and multiply by zero point eight for a realistic year-three usable number in a moderate climate, maybe zero point seven five if you're in a hot climate. And if that number still works for your needs, you're going to be satisfied. If it doesn't, you need a bigger system or better installation conditions.
That's a good rule of thumb. And insist on seeing the warranty terms before you sign. Not the summary, the actual terms. Look for the ambient temperature limits, the cycle count limits, and whether the capacity guarantee is at the cell or at the system output. Those three things will tell you more about what you're actually buying than any spec sheet.
Now: Hilbert's daily fun fact.
Hilbert: The national animal of Scotland is the unicorn. It has been since the twelfth century, when it was adopted as a symbol of purity and power in Scottish heraldry. Scotland is one of the only countries whose national animal is a mythological creature.
That explains a lot about Scotland, actually.
I have no follow-up to that.
So here's the forward-looking thought on home batteries. We're about to see a wave of second-generation systems hitting the market, solid-state batteries, improved LFP chemistries, and the manufacturers are going to make even bolder claims about longevity and capacity. The question is whether the transparency improves alongside the technology, or whether the gap between the spec sheet and the wall outlet stays exactly where it is.
The other question is whether third-party monitoring and independent testing becomes widespread enough that homeowners can actually verify what they're getting. Right now, you mostly have to trust the manufacturer's app and the manufacturer's data. That's not a great position to be in when you're trying to make a warranty claim.
Thanks to our producer Hilbert Flumingtop. This has been My Weird Prompts. Find us at myweirdprompts dot com or wherever you get your podcasts. We'll be back next time.
