Hacker News Reader

Comment by tedk-42 | Hacker News Reader

Honestly I'd rather a battery that's super reliable and NOT using any lithium ion inside it in my home.

As it doesn't need to move, it can sit there in an unused space and be really heavy. Lead Acid Batteries are a very good fit here (traditionally used as the 12V battery in cars).

In Australia you can buy a 12V 42AH (amp-hour) battery for ~$150AUD or $100USD. 20 of these will give you 10kWh @ 240VDC.

Couple this with a 6kW solar system and you'd be almost entirely free from grid electricity, especially in summer.

You can individually test out the batteries and swap out bad ones. The batteries themselves typically last >5years in a car and I imagine much longer at home where they aren't exposed to high engine temps and large continuous current discharges (engine cranking on start up).

View on HN

You might want to check LiFePO4 battery chemistry. Heavier than Li-Ion, but longevity is better, as well as safety. I have been told that you can hammer a nail through a cell and nothing spectacular happens.

(One) trouble with lead acid is that deep discharges are really bad for the longevity of the battery. So you can use only a small fraction of the capacity if you want to run regular cycles form it.

LiFePO4 is great - especially as an addition to solar systems currently using lead acid.

If you add a couple of LiFePO4 batteries before your lead acid (you should do some basic math where you have enough LiFePO4 to last about 1-2 days, depending on your location, without having to hit the lead) using a DC-DC charger (which does reduce efficiency - but that's ok) you increase the size of your pack, really help pull ALL the power out of your solar that you can, add a very safe battery (LiFePO4 is really safe compared to traditional lithium), and best of all, you really, really, really extend the lifespan of your lead.

I mean like, you probably will never be able to kill your lead doing this, for multiple reasons: primarily, you're just not using it that much, but you're also really keeping it topped off when you do use it.

The key is that lead acid isn't great at sucking up charge when it is 80-100% charged. So it is really hard to "catch up" charge with solar, even if you have a ton of it. Solar is not going to put out more than it draws. So if your batteries will only accept some X of AMPs to charge, that is all you're getting (well, you can use extra power for other stuff when this is happening, if you're set up that way, like powering appliances, so you can get some benefit still).

However, with LiFePO4 - that bad boy will take all the power you can throw at it (well, not technically, but for the purposes of a solar array, sure) and top off really quick. Then, when the sun is down, it charges your lead.

I think the benefits of this is more important for smaller arrays with fewer lead acid batteries, and where the LiFePO4 battery will be a single $500 100ah battery (rather than 10 of those), but I imagine even in large large setups this could really help.

LiFePO4 is amazing but I think the cheapest you can go right now by buying used server rack equipment is like $250 for $100ah? So 2.5x or more the price of lead, and you're comparing a used battery vs a new battery (and with the lead, $100 will probably get you a reputable battery, not some no-name used LiFePO4 battery).

If you are draining the LiFePO4 batteries faster than you can charge them and the lead batteries are charging the LiFePO4 batteries than wouldn't you have the same problem that will reduce the lifespan of the lead batteries?

You would probably need switch (battery isolator, Dc-DC charger, etc ) of some sort to isolate the two systems to fully eliminate this, but part of the answer is the Lifepo4 keeps its voltage higher through most of it's capacity than lead does. Thus the voltage of a Lifepo4 at ~20% is the same as a lead acid at ~100%.

This is true, but in my case I use a DC-DC charger that does not allow the lead to charge the lithium (I don't want this or want to think about it, even if it was beneficial, in my case the complexity is not worth it) - the only way the voltage can travel is from the lithium to the lead.

Okay and you discharge the lead first and the lead only fully drains when all of the batteries fully drain.

Yes this is correct - but by not really discharging the lead, you will keep it alive for essentially forever.

How does this switching black magic work. Is there a power interruption while it changes the source power bank?

No switching, the system looks like this:

Solar -> Charge Controller -> Lithium -> DC-DC Charger -> Lead Acid -> Load

The load is always pulled from the lead acid, but in reality if the load increases your DC-DC charger will just put out more juice (until it hits its limits) and not really drawn down the lead. If you max out the DC-DC charger yeah you'll start to draw from the lead but that is ok - that is what it is there for.

For a really large bank that you need to draw down lots of amps you may need to rework things, but likely lead acid doesn't work for you in those cases anyway. This works for banks where your load is like 50 amps or less. For my case, I have 1kAH including the lithium and I do not ever draw more than 20 amps. Works great.

Thanks for the extra detail. I need to go dust off my Kirchoff rule books. :-)

You probably know more than me already, honestly!

Sorry if I was unclear - it is the other way around. The LiFePO4 is before the Lead. So it is:

Solar -> Charge Controllers -> LiFePO4 -> DC-DC Charger -> Lead -> Load

In this configuration, yes, the lithium is always charging the Lead, and the DC-DC charger is creating heat. But the DC-DC charger is not two ways. So the lithium does not get charged.

But the benefits are above, and also you can make the lithium bank a higher voltage bank (so 24v, or 48v, or whatever) which has benefits for your solar: you can use smaller wire.

I am super interested in this, is there a grid tie solar box out there that supports heterogenous battery chemistry like you are describing?

Not sure if this is applicable?

https://www.bos-ag.com/products/le300/

Not sure - but very easy to make with off the shelf Victron products. I use their DC-DC charger and their charge controllers and have had great success so far.

This guy also has a lot of educational videos: https://www.youtube.com/c/WillProwse

I'm still thinking about this. What is the benefit of adding lead if you don't fully discharge the lead? Or do you?

If you are only using the lead until the LiFePO4 are dead then what is the benefit? Or do you have something to disconnect the batteries with they are at 50% (or what 55% to be safe) charge?

So you can increase the battery capacity by 50% of the lead you buy instead of 100% LiFePO4. Which means the lead needs to be half the price or less (excluding the cost of the DC-DC charger) for it to make sense.

Am I missing something?

Insurance.

You want your LiFePO4 to absorb the day to day brownouts/outages as well as sunlight variation and charge/discharge, and you want your lead acid there for when you get taken off the grid for 3 days and maybe have your solar ripped off your roof or partially degraded, once every ten years.

You also can afford to capture more energy faster and charge your lead acid slower - vs just dissipating the extra the lead acid can't use as heat.

>get taken off the grid for 3 days and maybe have your solar ripped off your roof or partially degraded, once every ten years.

Depending on location, the "taken off grid for 3 days and having your solar ripped off" could be from the same storm situation (hurricanes, tornados, etc)

In a setup where you only move between 80-100% charge for most of the time, dip to 50% maybe once a month, and to 10% once a year.

It's kind of like tiered storage on a file server: most capacity is rarely accessed but still valuable to have, so instead of using a lot of mediocre storage devices you use a few really good ones and a lot of really cheap ones. The good ones will stem most of the read load, and buffer writes to the slower storage if necessary.

> What is the benefit of adding lead if you don't fully discharge the lead? Or do you?

You do - but not always. If you only discharge the lead 1/10th of the time, your lead batteries will last incredibly long (if otherwise properly maintained).

Additionally, lead doesn't take up charge quickly. Putting in Lithium allows you to take full advantage of your solar. This can be mitigated in other ways, but this makes it "easy" and a not complex system with many charge controllers, many batteries, etc...

Additionally, it allows your input voltage to be much higher (from the solar) meaning you can have smaller wires (but amps are amps!).

> If you are only using the lead until the LiFePO4 are dead then what is the benefit? Or do you have something to disconnect the batteries with they are at 50% (or what 55% to be safe) charge?

I'm not sure what the question is, sorry. Yes you have equipment to make sure you are safely drawing from the lithium and safely charging the lead.

> So you can increase the battery capacity by 50% of the lead you buy instead of 100% LiFePO4. Which means the lead needs to be half the price or less (excluding the cost of the DC-DC charger) for it to make sense.

Well, if cost is not a problem, it would be better and a more simple system if you bought 100% LiFePO4. But that is really expensive.

This is nice idea that I watched some video about it few months ago, but will prowse (do you know him?) didn't like the idea as he prefers all lithium (he thinks it's cheap enough), I'm living in 3rd world country with a LOT of power outages (basically no power), I have a solar array with tubular lead acid but I'm soon switching to LiFePo4 when my package come from China, I feel tempted to implement your idea instead of selling my acid batteries. Do you think I can use MPPT controller instead of DC-to-DC charger? Technically it works like that I think

Yeah he knows way more than I do - and I agree that all lithium is probably the best way to go if you can afford it and are starting from scratch.

In my case, similar to yours, I had a ton of lead already - replacing all of it would have cost a lot of money, and would have been wasteful as the lead was only a few years old.

> Do you think I can use MPPT controller instead of DC-to-DC charger?

Only two issues I can think of, but definitely do your own research: 1) unsure if you can set up something for the MPPT to cut off charge if your source battery drops below a certain voltage, which you can do with at least some DC-DC chargers, and 2) it is common for MPPT charge controllers to lack the capability to boost voltage, just convert it lower.

#2 can be mitigated by setting up your lithium bank to provide higher current, which has its own dangers, but has added benefits as well such as requiring smaller diameter wire.

#1 is a serious concern and can also give you safety problems outside of maybe killing your lithium - so please make sure you're being safe and doing your research. maybe ask in diysolarforum.com ? I haven't posted there myself but have gotten great info from threads there.

Thanks for the advice I will consider posting a question on the forum when the package arrive. Will's opinion was a comment on this video btw, I can see now it's not totally negative though [1] For the problem #1 I think the BMS can solve the issue as I already ordered a JBD BMS so it can cut off whenever it reaches the critical voltage. For #2 you're right I think I should check, my MPPT is good one (Epever) but not sure it could lift the voltage properly from such a narrow difference, unless maybe I use higher voltage (such as 48) than my pack 's voltage (24v) so it will be guaranteed to only reduce the voltage to the lead acid levels. Thanks so much!

-- 1: https://youtu.be/tAuPfgZgXec

Yes - I have watched this video. Emily & Clark have a different setup, as he uses a different device (or even, no device) to connect the two different banks (lithium and lead acid).

I didn't see Will's comment tho, thank you for sending that, very thought provoking! The comment by him "Trickle charging lead acid all night with lithium seems very inefficient" is 100% true, but in my case I consider this effect a bonus b/c the lithium acts like a capacitor for the solar array (quickly charges, then slowly charges the lead!)

You're right, I think my situation is similar to yours. My solar array is not so big (for now at least) so I have lot of power in the noon but less in other times, so it will be very helpful to have a lithium pack that absorb all that power quickly and then feed it to the lazy lead pack over the next few hours.

yes, same - I only have 400W of solar!

I have 1540W and it's enough to cover more than 16hrs of power outages (light fridge, fans etc...), but I'm planning to expand as things are going wild in the near future probably. I'm afraid at the current rate of wars and energy insecurity even some rich countries will face the issues we're facing and we will be even worse!

I used to have an Epever MPPT. Really nice unit. However, as I recall, the manual emphasized that you should only use solar panels for the input. This is because when the battery is full the MPPT will stop charging by shorting the input. This is fine when the input is solar. But if you are doing LiFePO4 -> MPPT -> lead, it's going to short out your LiFePO4 battery when the lead battery is full, which would be very very bad.

I checked the manual and wasn't able to find this warning, but you might be correct. Thanks for heads up I didn't think of this scenario!

Woah, you can get a 10kWH LiFePO4 Powerwall for $2600. I didn't realize heavy battery tech is that far ahead. Seems rather compact too. That just seems entirely reasonable to install in your garage.

This is why I'm always mystified by the hoards of people that say that batteries will never be a significant grid resource. They are getting incredibly cheap already, even as our demand is exceeding supply.

There's some great modeling by Christopher Clack about how deploying tons of small storage and solar on the grid edge (at homes and businesses, next to the meter), and doing it right now, will enable far far more penetration of utility scale solar later.

Because, contra utility talking points, distributed solar and storage is actually a massive grid asset that lessens the transmission requirements and greatly lowers the overall cost of our electricity system.

We have the technology to get to 80%-90% renewable energy today, and at today's prices it will be cheaper than our current system. And by the time we get to 80-90% renewable power, other tech will have advanced far enough to go the rest of the way.

We just need to reshape regulations and markets so that the cheapest grid can be built, and that grid will be carbon free, and cause massive amounts of wealth generation. The only losers in this transition will be the corporations that fail to make the right bets on the future.

With RV Lead Acid Batteries, you typically can only discharge half of the battery safely. So a 100Ah battery really gets you 50Ah. With the LifePo4 batteries (also much less fire hazard than regular li-ion), you can regularly discharge >90% of the capacity.

>(One) trouble with lead acid is that deep discharges are really bad for the longevity of the battery.

Let me introduce you to the Optima Blue Top[0] marine battery which is designed specifically for draining to 0 and recovering to full charge. Doing this to a typical SLA battery found in cars is definitely not recommended. These are commonly used in fishing boats to run trolling motors and pumps in the boat's live wells. Being the hacker nerd type, I built a DIT cart around these with an inverter to power my gear in remote locations without needing puttputts.

[0]https://www.optimabatteries.com/optima-product-categories/ba...

Are the Blues much different from the Yellow Tops?

I've been running Yellows in my Jeep due to the heavy electrical load from the winch--I've nearly drained the battery a few times having to winch myself out with the motor off--without issue.

From a quick glance, they seem to be the same construction?

It's been a long while since I looked at the differences. I know the red top are typical car use for starting with high surge loads. these are not recommended for deep drain usage.

There are 2 blue tops, and just realized the one I linked previosly is not the same version (dark gray) as the one I have being the light gray.

Essentially, the differences are in use cases. Lots of short but heavy loads vs long sustained loads. Looks like the blue tops are still Optima's best "drain to 0 frequently" type of battery where the yellows look like an occassional drain to 0 is survivable but not the recommended use case. I had forgotten the differences after making the decision years ago. Just looked them up again for a quick refresher

> The batteries themselves typically last >5years in a car

In a home you will be using up a good chunk of the capacity every night. Lead acid batteries don't last many cycles[1], and I suspect you would be replacing them every 18 months if you powered your house off them.

Boats often have a solar+lead acid setup, and the owners of such systems typically have to make lifestyle changes to adapt around the limitations of the power system. Things like running only one appliance at a time, only starting the washing machine if the batteries are over 50%, never use any electric heaters, etc.

[1]: https://batteryuniversity.com/article/bu-804-how-to-prolong-...

This, op seem to think cycles, depth of discharge aren't important

What is your reason for skepticism of lithium batteries? The Australian megapack that caught fire during debugging? Hoverboards on airplanes?

I ask because I have a powerwall under the floor where i'm sitting right now and zero hesitation about it.

Indeed, with my ~6.5kw nominal solar, I am essentially net neutral.

The car in my driveway has a lithium battery 5x the size and hits 50 degrees during this heat wave. It also charges at about 3x the rate.

So the relatively cool crawlspace seems like a totally normal place to discharge at 5kw?

If you go with another chemistry, I think you'll find that cycling from 100% to 0-20% daily will not last for many years..

> 20 of these will give you 10kWh

Not really - lead acid batteries are limited to only using about 50% of their capacity, go any deeper more than a handful of times and you'll pretty much destroy it.

No, they're not. It depends on what battery you pick, and if you go with a deep cycle battery designed for that sort of storage use, you typically get charts of cycle life down to 80% DoD or so, and still have "hundreds of cycles" there on anything competent. With thousands of cycles for shallower cycling that you'd see on a daily basis.

Please go find a good, modern lead acid datasheet and familiarize yourself with it, because I'm tired of people who've clearly not been around anything from the last decade or two parroting this line endlessly when it comes to lithium and lead.

Here's the datasheet for a GC2 form factor Trojan T105-RE from about a decade ago (I've got a 6 year old pack in my office): https://www.trojanbattery.com/pdf/datasheets/T105RE_TrojanRE...

Notice that it's rated for about 700 cycles, to 100% DoD. And about 4000 cycles at 20% DoD.

Lead has many advantages for stationary storage, and they're not the same batteries you found 40 years ago.

Most of the time I've read about DIY systems, deep cycle marine batteries are preferred for this reason.

The apples to apples comparison of capacity kWh is misleading. The cheaper lead acid should not be discharged below 50% regularly, cutting the useful kWh in half for regular use.

When I did the math for my RV on Lifepo4 vs lead acid, the cost per usable kWh was higher unless you looked at the expected lifetime replacement costs, then it was even. I haven't followed battery prices too closely but my impression is that Lifepo4 has dropped in price ~30-50% since then so is probably now cheaper for usable kWh×cycle than lead acid but still might have a higher upfront cost.

The supported tempurature ranges, space contraints (venting, size, weight) and other factors all seem to favor Lifepo4.

That said, if your house battery is a backup for power outages and not intended to power your house every night, then lead acid may still be the way to go because you don't care about kWh not kWh×cycle.

Edit: replaced * with × to fix formatting

Is there any data about battery safety, like NiMH vs LiFePO4 vs lead acid [add your favorite chemistry here]?

I understand LiFePO4 batteries to be exceedingly safe. They don't overheat or catch fire even if punctured.

Lead acid batteries, well, contain lead, which, as a father, is not something I want in our house. It can also produce hydrogen and oxygen, potentially causing an explosion.

Lead acid batteries can also produce hydrogen sulfide, which occasionally kills boat occupants using them.

That is basically what a forklift battery is. You can pick those up for the scrap cost of the lead as many places either recycle them on a timeline and not when the battery is dead, or recycle them when a cell or two needs replacing. I've seen forklifts with 20-year-old batteries - being in a heated climate, never being deep cycled, and the purity of the lead are all factors in longevity.

I would still want to house that power pack in a brick box on the street. If only to simplify any emergency response and save my house.

You can buy rack mount lipo batteries and buy a climate controlled nema 4x rack and put it all away from your house.

I have this, just haven’t installed it yet. $5500 for a half rack.

Lithium Titanate (LTO) is the only lithium based home battery I would feel comfortable with but same sentiment on safety. The chance of thermal runaway is low but results are catastrophic.

I'm not familiar with LTO, what advantages does it have over LiFePO4?

It's basically LiFePO4 but even more so: really high charge (10C! and it doesn't even get warm to the touch) and discharge rates, even harder to damage to the point of failure, and really long lifetime (many, many full charge/discharge cycles). Slightly worse energy density. Main downside is it's pretty expensive and niche still. I wouldn't necessarily consider it optimal for solar power storage, where it really shines is high duty cycle high power operation like electric busses, because you can fast-charge it multiple times a day for years.

Spot on. The cost is definitely the main issue IMO. I think they make for great option for home backup/offgrid use mainly due to their crazy high cycles like you pointed out. An Australian company (Zenaji) sells a solution with a warranty on 22,000 cycles. Again, super expensive (~$3k AUD per 1.95kWh) but IMO a superior tech other than density (not a huge difference but definitely not as good as LiFePO4/NMC), which doesn't matter as much in home/commercial stationary backup scenarios.

Yes I've seen very little about lead acid batteries for solar storage. Considering that home installments are almost never weight constrains and often size is not a big issue either, they seem like a perfect solution. I wonder if there is some important bit of information missing that makes them unsuitable?

The cheap lead-acid batteries you find in your car don't handle deep discharge very well, let it go empty a couple times and you have to replace it (at least they can be recycled easily). There are variants that can handle that better, but they cost easily twice as much. If you charge them too much they start venting hydrogen, which could build up to dangerous levels in buildings. I think that was discussed as a cause in the OVH datacenter fire.

So they do have some problems, but still almost all battery backup is done with lead acid (with lithium-iron catching up as a good replacement). I think most of these li-ion backup batteries are more about using surplus production when you scale your battery production faster than vehicle sales.

Lead acid batteries are commonly used for solar storage. They require more physical maintenance than lithium ion, and they have more specific charging requirements.

There is really nothing "perfect" about lead acid.

They let out fire-hazard gases when they charge. You need to maintain their internal water levels. They tend to spew acid on their surfaces. Their voltage droops as you use them. They can't tolerate deep discharge at all, and in general wear out quite quickly when heavily used.

LiFePO4 is a huge improvement over lead acid.

Having 20 lead batteries in your basement also adds a lot of prepper flair and scares away those annoying fire protection guys too. It is just 10kWh regardless of voltage aside from small conversion losses here and there.

Provided you have a computer that uses 500W, a monitor that uses 100W plus your internet router of 20W, you can run it for a bit over 16 hours. Just answering the important questions.

Although these type of battery also wants to be handled correctly. Don't discharge completely and yet also do not charge them fully. You also have to know about the hundreds of other partially true myths about lead batteries. Their most prominent positive property is indeed the price and not too much else.

Backyard sheds seem like the best place for battery packs this large.

However, I think a lot of people with electric cars and attached garages don't think twice about it.

I guess that rules out 99% of city dwellers then.

Are you sure your lease doesn't rule it out anyway?

Where would a city dweller put a huge battery pack anyway? Maybe ask your landlord to put one in the basement for the whole building? Obviously he won't; it doesn't provide any benefit to him.

> Honestly I'd rather a battery that's super reliable and NOT using any lithium ion inside it in my home. As it doesn't need to move, it can sit there in an unused space and be really heavy.

It sounds to me like you're asking for a nickel-iron battery.

You want about 20kWh at a minimum to be self sufficient. We have a 10kWh Tesla battery, and it only lasts until the evening (saving 20%) for grid outage backup.

For your case.

Because this really depends on climate and the house.

In our case, 100 kWh would be required to cover one day normal power requirements in the worst case. Median would be about 50 kWh.

I could probably stretch 100 kWh for max 3 days (worst case) by adjusting temperature, using water minimally and unplugging everything that's not absolutely required.

There are also households who could stretch 8.7 kWh for several days. And houses that require significantly more power to maintain liveable conditions.

whats the opposite of a humble brag

My point is one size doesn't fit all.

Some places on this planet can get very cold or very hot. If you're running a [ground source] heat pump as your primary heat/cooling source, even with high COP, on extreme days the system can draw massive amounts of power.

You know, those days when losing power could truly hurt. Like destroy water pipes (including underfloor heating pipes) type hurt. Or make the house unlivable hot.

No bragging (I have no idea how someone could read it like that!) or anything. Just facts. There's more variability than you might think.

That's probably at "full load" for your house though, no? If you're looking at an extended outage, you can alter your use patterns.

I have a 5kW gasoline generator that most certainly cannot run the whole house, but in emergency situations, I turn off most things at the breaker box. We can run one AC unit to cool part of the house, then shut down that, and power up the kitchen for dinner, etc, etc.

Reply on news.ycombinator.com