LA versus lithium battery sizing

jtdiesel65

What's the current consensus on battery sizing? I have about a 10KW/day load. Currently on 24v but moving to 48v.  I currently use

https://gb-battery.com/products/ols/products/new-12-125-17-battery-24-volt-1000ah

It's has positive plate growth and one cell that doesn't look too good. It's about 9 years old.

Pricing seems to still point to forklift batteries, but if I read correctly lithium may blur the math due to DoD.  But what are people doing with regard to DOD on forklifts and lithium batteries?  I've been charging at about 60-70% remaining, but have heard of some people bringing forklift batts down to 20% and others who charge at 80%.

.

I was looking at maybe getting 4 of these

https://gb-battery.com/products/ols/products/6-85-17-sb-slash-1-072-ah-solar-battery-module

which is 938 ah and about $7000.

Is there a lithium alternative that even comes close to the price?

also, how does the max output of lithium work with big startup loads like AC units?

thx

Photowhit

jtdiesel65 said:

....I've been charging at about 60-70% remaining, but have heard of some people bringing forklift batts down to 20% and others who charge at 80%. 

So... How are you charging your battery bank? A generator?

I hooked mine up to the solar electric system and let it go, charged every day.

Yes they say not to do 'short charging' with the battery not drawn down a good bit. I choose to believe that has to do with the intended purpose as a 'forklift battery', not wanting people to 'hot charge' between shifts. Since solar is a much more gradual charge, in general. I just ignore that bit...

jtdiesel65

Photowhit said:

jtdiesel65 said:

....I've been charging at about 60-70% remaining, but have heard of some people bringing forklift batts down to 20% and others who charge at 80%. 

So... How are you charging your battery bank? A generator?

I hooked mine up to the solar electric system and let it go, charged every day.

Yes they say not to do 'short charging' with the battery not drawn down a good bit. I choose to believe that has to do with the intended purpose as a 'forklift battery', not wanting people to 'hot charge' between shifts. Since solar is a much more gradual charge, in general. I just ignore that bit...

The controllers aren't limited. They charge when there is sun

The AGS 2 hr trigger is set at 24.2v and that seems to be the one that starts the generator now days. According to classics this is about 360AH of usage from a "full" battery.  That's about 8640 watts at 24v. This seems to jive with the combox data that says I'm using 8-12kw /day. But, 24.2 volts should be more in line with a 20% of charge remaining which would seem to indicate that the battery capacity has dropped to 450AH. Specs on it new are 1000ah at 6hr and 1576ah at 20hr.  Specific gravities of most cells are in ~1.2 area when the voltage is in the 24.2 area. There is one cell that reads low.  It does have positive plate growth. I think it's dying a quickish death at this point. Visually there is darkish sediment that you can see on the matt in each cell. I'm not sure at what point the plate growth would crack the plastic.

Anyway, I need to get a new battery. Lithium would be nice but the cost seems prohibitive.  If you want even 2 days of autonomy, that's getting into price of small car.

I was thinking of going lithium and just adding to the bank every few months. But then, manufacturers could discontinue the product at any time.

mcgivor

The cost of lithium (LiFePo4) largely depends on wether pre assembled modules are used or a DIY bank is assembled, the latter is roughly half the cost using top of the line cells. When compared to flooded lead acid the cost is roughly double using a direct capacity to capacity, but this comparison is misleading due to the depth of discharge allowable to achieve a reasonable cycle life expectancy, so it's comparing apples to oranges. When all factors are considered keeping within the safe limitations of each chemistry, the cost per kilowatt over the lifespan is roughly equal.

To explain this one needs to consider that in order to achieve a reasonable cycle count with LA, the DOD needs to be kept shallow, therefore a larger initial capacity is required to compensate, for example a 1000 Ah bank is really a 300 Ah, so it's either pay up front or deep discharge and replace frequently this article will help explain  https://www.powertechsystems.eu/home/tech-corner/lithium-ion-vs-lead-acid-cost-analysis/. Note the link is using 50% DOD for LA, 100% for LFP to achieve a direct comparison, using lower DOD in both chemistries would have different results but in the end the results would still favor LFP. 
https://www.fortresspower.com/how-to-calculate-the-energy-cost-of-different-battery-chemistries/

Also read the attached pdf, there are many comparisons which can be found which come to similar conclusions, so yes the upfront cost is higher but all things considered its lower.

Looking at your consumption of 360 Ah at 24V according to the Classic, assuming that is charging ~20% of that would be losses due to inefficiencies so 288 Ah would be more accurate. Using 300 Ah  as a baseline my personal opinion it would be best to have double that for LFP to allow for bad weather and to significantly increase cycle life expectancy. With 600 Ah that would have a nominal capacity of 14.4 kW, my cost for a DIY bank using 24 grade A CALB 200Ah  cells would be $ 3,240 (2020 price) excluding shipping, a BMS will cost anywhere between  $60 and $200 depending on features required. (A 48V bank would be 300 Ah same cost)

Building a DIY bank is really simple contrary to popular belief, LFP is far superior in every aspect except in temperatures near or below freezing, they are around 98% efficient, charging is extremely fast without a long absorption making generator charging highly efficient and no maintenance or corrosive emissions.

One additional piece of information omitted is, lead acid has a recycling value, I recovered 20%  scrap value on initial cost, this may vary depending on location.

mike_s

You'll find a lot of comments about only discharging FLA to 50%. Yes, that results in longer chronological and cycle life, but if you look into it more, discharging further doesn't change the lifetime cost/energy delivered by much. e.g. if you regularly go down to 10% and get 1/2 the cycles, your cost isn't much higher. A properly sized bank is only going to need to provide a deep discharge in unusual situations - long term overcast, higher than normal usage, etc. And, batteries age whether used or not, so that makes the difference even less. Li also offer fewer cycles with higher DoD, although the drop is a bit less.

Trojan shows DoD/cycle life in some of their documentation. 1900 cycles@50% vs 1000@80% and the curve is flattening (ratio is similar with different product lines). LI does offer more cycles overall. Except, for solar applications, 8000 cycles is 21 years. I'll wager that with daily solar cycling, the battery isn't going to last long enough to make that useful. If you're cycling multiple times per day, sure.

You can get 225Ah x 12V x 1200 cycles (3 years at 50% daily, but much more based on real world reports) FLA  for <$400 (Trojan T-105, assuming no core trade-in). Or you can buy a similar capacity Li for $2100 which might last thrice as long. Prices from our sponsors. I just picked a baseline. Scale it up as much as you want, the balance doesn't really change much. 
Most comparisons ignore the cost of money. Take that 80% savings of FLA vs Li, invest it (long term inflation adjusted SP500 averages 7%/year), and you'll have enough return to buy a new set every 3 years.

I've tried to use conservative figures, feel free to be more conservative. But it's hard to make a case for Li without extenuating circumstances. Lithium are fine if the requirements for size and weight are important (e.g. vehicles), I can't see much reason to pay the premium otherwise.

Dave Angelini

I know of more than a few Li horror stories. The one I can talk about (because he said I could) was a client in Canada who knew of the cold weather perils and killed a 22K$ bank. He was called away to a funeral, the genset failed, he lost heating, and could not get back for 2 weeks as the road was washed out. Keep in mind it is not like you get a warning, just one time is all it takes.

I really do not price check alot because I use what is appropriate for the design. On the back of a bar napkin I would write 3 dollars for LI, 2 dollars for AGM, and 1 dollar for flooded.

jtdiesel65

Thanks for the info.  I don't see a prepackaged LI solution at this point. Cost just doesn't work out.

I'm moving to 48v and will likely stick to LA this iteration.

thanks for the comparisons.

mcgivor

Dave Angelini said:

I know of more than a few Li horror stories. The one I can talk about (because he said I could) was a client in Canada who knew of the cold weather perils and killed a 22K$ bank. He was called away to a funeral, the genset failed, he lost heating, and could not get back for 2 weeks as the road was washed out. Keep in mind it is not like you get a warning, just one time is all it takes.

I really do not price check alot because I use what is appropriate for the design. On the back of a bar napkin I would write 3 dollars for LI, 2 dollars for AGM, and 1 dollar for flooded.

As a client why didn't you suggest an appropriate BMS protection system to prevent the loss by disconnecting loads when discharged below a safe value and disconnecting the charging source when below 5°C. Designing for worst case scenario's with redundancy, is a prudent thing to consider, at least that's what I would think for mysel, or clients, if I had them.

Dave Angelini

You do not know what you are talking about and you do not know the whole story.
Probably the reason you would never have clients also. Please do not quote me anymore!

mcgivor

Dave Angelini said:

You do not know what you are talking about and you do not know the whole story.
Probably the reason you would never have clients also. Please do not quote me anymore!

Just asking a simple question promts a judgmental insulting retort?

Don't judge me by making suggestions I don't know what I'm talking about, you only offered part of a story, so naturally I would not know the whole story, hense the question.

Secondly I don't need clients and certainly wouldn't use a forum to promote myself if I did.

Dave Angelini

The guys parents died in a head on. He knew the perils as I said. His luck ran out ! What would you have me do? Rub it is and say I told you so.

 I am helping JT here not to make a stupid mistake.

jtdiesel65

Other than cold, what are the main causes of a failure for LFP? BMS going bad? If I understand correctly, the proper set up is roughly

AGS trigger volts> inverter shutdown volts> BMS low cut off volts and

Inverter/charge controller charge absorb volts < BMS overcharge protection volts

IIUC, LFP state should bounce between 10% and 90% of the capacity and the BMS shouldn't be doing much other than monitoring individual cells and temp. A BMS low or high cutoff should never happen if the equipment is configured correctly.

It looks like DIY LFP, is slightly more money than FLA. The process of putting together a bank of LFP cells doesn't appear to be difficult.  I would think lack of long absorb for the LFP would save propane in winter and maybe free up some watts when charging from PV. 

The lack of maintenance for the LFP is nice, although the forklifts have some watering system options.  The FLA is attractive because everyone knows FLA. If you have to get someone to troubleshoot a problem, it might be better. The other thing is size/weight. My current 24v forklift battery is 2000 lbs. A new Forklift battery will have to be either the company's new solar options with removable cells or 12 v strings so weight is manageable. Weight/size isn't a concern for LFP.

Dave Angelini

LFP are great as long as you can keep them warm and never charge them below 35F. Only you can make that call in Vermont I think right?.

They are killed in large power systems when they are comprised of small batteries and the minimum number of batts is not there.
They are charged based on Soc not volts in large systems. XWP is built for this and the gateway can get data from a well designed LFP battery BMS.

If you have insurance on your home, you could lose in a claim if your LFP battery does not have listings. Insurance agents have seen the YouTube of battery fires, even though most of that is NMC lithium. Old flooded/AGM technology gets grandfathered in. LFP does not! You should inquire with your insurance company!

jtdiesel65

Well, I'm not too worried about temps. They would be in a heated building. This has a simple propane no electricity required heater with an electric backup on a thermostat if if fails.  I would certainly use a BMS with a low temp cut out.

If I were to go LFP, it would be minimally 560 AH to start and would likely double it to 1120AH. I was reading the XWP li ion solution guide and it looks like 560AH (~27KWH ) would work.

I already have a gateway, battery monitor, and AGS. Any idea which BMS would work with the gateway/battery monitor?

thanks

mcgivor

More like which battery system with an integral BMS will work, where communication is required between the battery and balance components, Schneider has partnered with battery system manufacturers to share communication information to ensure a safe system for those who don't want to figure it out for themselves, admittedly is a sound decision for a corporate player, to ensure all testing is done prior to release to the public.

Conversely there are those who don't require the added expense associated with the research and development but rather source equipment based on requirements, it's a personal decision. If one is unwilling, or lacks the ability to construct a DIY system, then perhaps relying on a proven system is a better solution. This doesn't necessarily mean one is inherently better than another if all safeguardsare considered, however I digress.

mike95490

The big failing I see in home brew, using "just any" BMS, is the ability for the charger, to keep dumping amps into a bank that has 1 full cell.  It's balance or bypass board is usually only able to manage shunting an amp or 2 around the cell, while the charger is dumping 20A into it.

mcgivor

mike95490 said:

The big failing I see in home brew, using "just any" BMS, is the ability for the charger, to keep dumping amps into a bank that has 1 full cell.  It's balance or bypass board is usually only able to manage shunting an amp or 2 around the cell, while the charger is dumping 20A into it.

The BMS monitors individual cell voltages disconnecting charging to the battery as a whole when an individual cells voltage reaches i maximum threshold, cell balance occurs at fully charged state where current flow is almost or at zero. Depending upon the BMS used, this may result in disconnect of the load as well as charging, as in the case of a single port BMS, a dual port type has the ability to individually disconnect one or the other depending on state of charge when the first cell reaches maximum or minimum thresholds.

Since the typical charging profiles used don't venture anywhere near the fully charged state this is generally a non issue but remains there as a safeguard should something go wrong with an individual cell. Programming the load to shut down well above the minimum cell voltage threshold prevents overdischarge, this is important to those using a non programmable inverter which typically have a fixed value tailored to lead acid, which is far too low for lithium, in such cases an auxillary voltage sensing method would be required to shed the load.

Many such concerns about lithium steer users to abandon the choice, once studied however, it becomes clear that these concerns have solutions. Having others do the research and development  possibly with a UL listing for insurance purposes along with a warranty, comes at a premium, so perhaps a DIY bank isn't for everyone considering the chemistry.

wild01

If you go the DIY route, right now aluminum cased lifepo4 cells drop shipped out of China right now are dirt cheap. THERE MAY BE A REASON FOR THIS They might not be A grade matched cells. Some people suspect they are B grade cells that have failed inspection. I personally find this to be unlikely due to the sheer volume of cells being sold. I think it is far more likely they are simply ungraded cells strait off the factory line w/o the complex and time consuming inspection and grading processes. Long and short of it, buyer beware. Having said that... many people are getting very good results from these cells. There are review videos all over youtube. It seems rare that they are more than 5% off their rated capacity and many people have been running them for over a year with good results. On Diysolarforum there is a group buy thread where one user has set up direct drop shipping. it appears most participants have been very pleased with the service. I was just quoted about 3k for 32 272ah lishen cells. that would make a 48v 544ah battery. 2 100a bms's with low temp cutoff would run $160-250. that would give you around 10kw charging and discharging support.

To me at this price point it's worth the risk. Downsides-no warantee, literally comes on a slow boat from China-1-2 months, upside cheaper than lead acid, much less maintenance, no damage from partial recharges, ie less gen run time. I'm gonna test the waters with a 272ah 24v bank in my rv first, if it performs well, I'm converting everything.

Raj174

Why two BMS's? Will there be two batteries or one 48V battery? If the 32 cells are 2P16S then only one BMS of proper amperage is necessary. That's the way I configured my recent purchase. Looks like this.

mike95490

Alligator clips on tarnished bus bars are not functional connections for a BMS.   And it's not a BMS except in the sales pitch. That's a battery voltage monitor with hi & lo cutoffs (maybe both if you are lucky). 
  What's the plan when the battery disconnects ?  Does your charge controller become unloaded and float up to your PV array Voc ?

 I see no thermal sensors, air gaps and no thermal separators in the pack, everything is all cuddled together, with the cells in the middle, likely 5 degrees warmer than the outer. 

Raj174

Thanks Mike, 

You make a valid point concerning the alligator clips. They are actually temporary and will soon be replaced. As for the rest, you should know that when you make comments like this, people that have actually built and managed a DIY LFP bank know immediately that you have never built and managed one yourself. And no matter how well read you are, that still makes you an arm chair critic.

Imagine me, living in Florida, telling you what it's like to live in California, even though I've never been there, just listened to what people say, and read about it in a book. You want people to take you seriously about this matter? Make the investment, build the battery, and get some experience. You know what I'm saying is true.

Making this statement is really burdensome to me because I have the utmost respect for you due to your knowledge and long years experience in the field of solar power, not to mention you living off grid. In fact, I have learned a quite lot from your posts over the last 5 or 6 years and I thank you for that. 

Rick

Dave Angelini

You have been great Rick also! I just think that when the store here sells a proper "kit" for someone to DIY, I will then start being interested. Saving money is down the list for me as a business person. The costs get passed on. In return, the client gets a safe UL listed system that can be insured by a homeowners policy. My only comment on what Mike wrote is the combustible surfaces (plywood) would probably never pass an inspection by either a building department or an insurance inspection. The wiring needs more protection from perils.

I do not like the 4 or 5 times I have been hired to guess what started the fire. Not fun work and none of them were my clients.

mike95490

Sorry Rick.  Even in the armchair, there are clues to BMS systems.  A single volt sensor across a pack of cells, is the barest low end BMS no matter what the salesman says.   And while Li doesn't care about voltage while at different temps, warmer cells don't last as long a cooler cells.   having a couple temp sensors scattered around on the bank is never a bad idea. 

Even with NiFe, I log & monitor volts and temps daily.   At some point one of the cells is bound to go south and i want all the notice I can get about it.

Raj174

Well, we agree about one thing. All cells do die eventually.

mcgivor

The Chargery, as seen in the image, is not a simple cell voltage monitor it is a true BMS capable of charge / discharge disconnect based on a multitude of fault conditions, it has two temperature sensors and a shunt based capacity readout. Although primarily designed for EV applications it can be utilized for offgrid, athough not the best choice in my opinion, due to its external contactor requirements. 

Having ill informed misleading comments constantly being made is frustrating to those who have actual hands experience and are attempting to share information with others who are in the beginning stage. When contemplating LFP I was discouraged at first but discovered through extensive study, along with personal dialog of existing users, that my concerns were easily overcome.

There are those who would rather build a DIY system rather than paying double for a manufactured battery system and or want the challenge of doing so. Even a DIY system could be approved to comply for insurance purposes if inspected by a listed UL preferred partner, if required.

wild01

Raj174 said:

Why two BMS's? Will there be two batteries or one 48V battery? If the 32 cells are 2P16S then only one BMS of proper amperage is necessary. That's the way I configured my recent purchase. Looks like this.

the reason I'm thinking 2 bms's is I haven't found a true 200a bms I like. I know 16s2p is not as ideal as 2p16s but it should function acceptably. another option is to build a stand alone low temp disconnect on and connect it to my incoming solar wires via a no relay (so it only uses power when I have incoming solar) I live in the frozen hell of Hoth so a low temp cutoff is a must. for the main setup I'm thinking of building the pack in a deep freezer body with a thermostatically controlled heater circuit and greenhouse wax motor vents on the lid. idea is it heats to 50f and opens the lid if the box goes above 70f.

mcgivor

When sourcing a suitable BMS what I did was make contact with manufacturers outlining the requirements of my system, cell topography,  maximum charging/current, being solar and so forth. The one I settled on is pictured below, the prime reasons were it has separate ports for charging and dischargeing, which allows charging when the low voltage disconnect has activated, unlike a single port type which disconnects both . The screw terminals are sized appropriately which allows heavy conductors to be directly connected, it however has no remote connectivity nor low temperature charging disconnect, both of which I have no need for being I'm in the tropics and use a secondary passive BMS, the Chargery as pictured above, to provide a display and backup protection.

Not sure what exactly you're looking for in term of criteria other than 200A but perhaps they have something to fit your needs, or not, but it's worth inquiring with manufacturers, here is a link   http://www.lws-pcm.com/en/produce.asp?Page=6  inquire in the contact us menu, I have no affiliation with the company, just a satisfied customer.

Horsefly

Just FYI, I have found at least one BMS (Overkill Solar, which rebrands a Chinese brand) has a single (common) port, but does separate the charging disconnect from the discharging disconnect. I had to go back and brush up on my FET knowledge, but you can use two FETs and two bypass diodes to have the current blocked in one direction or the other, or to allow both. I verified the Overkill BMS works that way by forcing a low temp disconnect on the charging, and immediately turning on a load. The charging did stop, but the load powered up immediately.

My guess is that there are others that do the separate charge / discharge disconnect on the same common port.  It looks like the newer Chargery BMS (with the solid state relay disconnect rather than the contactor) SSR has separate inputs for the charge and discharge controls from the main BMS, and the documentation sounds like it controls them separately. I.e., the charge can be disabled while the discharge is enabled, or vice versa.

wild01

The jdb bms is the one I've been looking at. (Same one overkill rebrands) only thing I don't really like about it is the 100a max. But for the price is the best one I've found. Well reviewed and well liked.

MichaelK

I've often heard the claim that Li batteries can be drained down to 0%, but is that really true?  I've seen recommendations from BattleBorn that their battery's lives are severely shorten once drained past 80% (I read a 5 fold reduction in life expectancy), and I've also read that Li batteries don't like routine charging all the way to 100%, and 90% is a recommended target.

So, in the real world, if people should be managing their Li batteries between 20-90% charge, it seems you are comparing a 70% battery to a traditional 50% lead-acid battery.  Is that conclusion off-base?  It appears to me that totally draining a Li battery shortens it life as much as 100% draining shortens lead life?

BB.

Yep Michael, That is what I have seen too... If you are going to pay 3x as much for a battery to last 3x longer--Following best practices is important.

-Bill