Taking accurate voltage readings of batteries?

ligwyd

I have 3 strings of 12 volt batteries in my battery bank. Four batteries in each string. 48 volt battery bank.

Is it even possible to get an accurate voltage reading from any point in the battery bank, given that all batteries are connected? Do you have to disconnect all the batteries to get true voltage readings from each battery?

I am also regularly checking specific gravity on all 72 cells with my hydro-volt hydrometer.

Also, I have a Schneider battery monitor that monitors 2 midpoint voltages (wish it had one more mid-point voltage tap wire so I could monitor the third string without using my volt meter).

While the battery monitor does display each of the two tapped midpoint voltages, can you actually trust that reading to be accurate, given that again, the batteries are all connected and as such, are having their voltage readings affected by the other batteries connected.

mcgivor

The individual batteries within strings can be measured for voltage without disconnecting each individual to establish if there are variations, but it's just a quick check method between SG readings to gauge how things are. More important is current  ballance between strings during charging / dischargeing, this is where a DC clamp on ammeter is an important tool as voltage itself can be misleading. Taking  SG readings of cells is the only way, in my opinion to  really know the health of a 12V monoblock because often there can be one undercharged cell and five overcharged, but voltage looks perfect.

Estragon

It seems to me the strings will tend to converge in voltage over time while resting, and diverge when under significant charge or discharge.  As such, individual battery voltages at rest are of limited use.

As Mcgivor notes, current balance while charging is the important metric.  To some extent though, this could be implied indirectly with individual voltage reading by seeing some rising faster in voltage than others.  The fast ones end up overcharged, and the slow ones undercharged.  Over time, the slow ones get slower, etc. 

I've sometimes used individual battery  voltages at around the start of absorb as a quick and dirty check between SG readings to see if the SG check should be done sooner than planned.

ligwyd

Any suggestions on a good clamp on DC Amp meter? I guess this is where individual shunts on each string would be a nice, provided that the equip. can monitor all three strings, But if you have one shunt as in my system, a clamp on amp meter will work fine too.

I really liked how this guy set up his battery compartment using three separate shunts. Much smaller footprint of course since he is using Lithium batteries. Just like the cleanness of the setup.

https://www.youtube.com/watch?v=onbndjtKdsA

In the future I will try to design FLA battery banks with one string only. Would also love to build a system soon and try two of these 48V 130AH lithium batteries in parallel. Albeit more investment up front.....and gotta keep em warm....(above freezing)

https://www.solar-electric.com/discover-energy-aes-6-6kwhr-48vdc-lithium-ion-battery.html

Also, I charged and equalized each string separately, and took voltage readings (and S.G. readings) before I hooked the strings back up  to the common positive and negative buss bars, and as has been mentioned here, all battery voltages come relatively in line with each other when not charging or discharging.

I think I may have a bad batch of batteries. I bought 8 first and then 4 more 3 months later (as an after thought I had decided to expand the battery bank capacity - first 8 not yet put into service - only maintained). Funny thing is, it is the newer batteries that are acting up and the older ones are working perfect. The newer 4 have lower S.G. than the 8 older batteries even after several one hour equalization's, and a few of the newer batteries have a couple cells that will not reach 1,255 S.G. or higher when nearly all of the other cells in the older batteries easily rise 1,270 S.G. and higher after one 50% discharge and recharge while the cells in the newer batteries are all lagging around 1,245 S.G. or lower.

Also, when I look into the suspected problem cells with a flashlight, you can physically see black deposits on the lead plate material visible through the fill caps - sulfation. All the batteries have been treated the exact same since I got them. I will be talking to the supplier, as they are still under warranty, Monday morning to present my case. We'll see how that goes?

Estragon

I've read that FLA batteries take a few dozen cycles to "form the plates" and reach full effective capacity.  I'm not sure exactly what's involved physically/chemically in that process, but it stands to reason that if true, the newer/lower capacity batteries will discharge more deeply relative to older in parallel, and may not be fully recharged with the older battery behavior masking this.

Under the circumstances, I'd be inclined to separate the new ones and go through a full charge/EQ on them alone.  If your loads allow, running on just the new ones for a few days might help too.

I got a usable DC clamp meter from amazon for ~$50.  @bb had links to some in recent thread(s).

BB.

A couple Amazon links to start with DC current clamp meters:

https://www.amazon.com/gp/product/B019CY4FB4 ($105)
https://www.amazon.com/gp/product/B07546L9RT ($40)

The "cheap one" may come with a Chinese only manual... So you may have to play around to learn how it works (or ask a few questions here).

-Bill

ligwyd

Thanks for the links. Re: my batteries, I like the theory, hopefully it holds true.

Still not sure why there is black spots on the lead bars when looking in through the fill caps? And the cells with the most black spots are the ones than will not reach full S.G. readings? Also, the older batteries were never put into service. Initially I charged all batteries at a minimum every three months, or whenever the first battery hit 12.3V (20% D.O.D.). Whichever cam first. So the first 8 batteries only have one more "top up charge on them, than the other newer 4 batteries. Still a little puzzling. Lithium is looking better all the time. Maybe exchanging certain problem for others? Experience will tell. So we'll take er one step at a time.

Appreciate you guys. Thanks

John

mcgivor

Regarding the video in post #4

When building a LFP battery the topography is opposite to how lead acid would be, all cells are first arranged in parallel to achieve the desired capacity then series to to get the required  nominal voltage, this means there are no parrallel batteries, just parallel cells. Using multiple  unitized pre built batteries in parallel may have advantages that I'm unaware of, other than redundancy perhaps , but having a single BMS looking after the entire bank definitely sounds simpler, to me at least. 

Understandably some don't want to build a DIY LFP bank, prefering manufactured solutions for personal reasons, but building your own is not difficult and once operational is definitely much easier to maintain than a LA bank. With a good BMS there is protection against many potential problems which may occur, things a lead acid bank would definitely benifit from but typically do not, which is why there is so much discussion regarding low cell capacity, over changing, over discharge etcetera.

ligwyd

Would love to learn more about building my own LFP batteries down the road. On lots of learning curves right now but will look into this more in the near future. Compared to say one of the new Discover AES 6.6 kwH batteries, how much more cost efficient can you build a comparable battery for? (not including your time - as I look at it as a priceless learning experience) .

I am beginning today, to discharge to 50% D.O.D., over approx 20 hours, and then recharging each of my three strings of batteries separately and I will record all S.G. after discharge recharge. This should give me a clearer idea of what is going on with my batteries. Then if I still am having concerns I'll approach the supplier with more than enough evidence to support my claim. 

Also, I have done so for now, but is it necessary, to disconnect, from the positive and negative bus bars, both the positive and negative battery parallel cables, or is just dis-connecting the positive battery cables alright? (when discharging/ charging each of the three string separately)

Estragon

If I'm understanding the question correctly, disconnecting the positive from a battery ( or string) to the parallel connection (bussbar) would open the circuit and isolate that battery.  Obviously, if you have a loose positive still connected to the battery, you'll want to be pretty careful in where that loose end goes

mcgivor

ligwyd said:

Would love to learn more about building my own LFP batteries down the road. On lots of learning curves right now but will look into this more in the near future. Compared to say one of the new Discover AES 6.6 kwH batteries, how much more cost efficient can you build a comparable battery for? (not including your time - as I look at it as a priceless learning experience) .

I am beginning today, to discharge to 50% D.O.D., over approx 20 hours, and then recharging each of my three strings of batteries separately and I will record all S.G. after discharge recharge. This should give me a clearer idea of what is going on with my batteries. Then if I still am having concerns I'll approach the supplier with more than enough evidence to support my claim. 

Also, I have done so for now, but is it necessary, to disconnect, from the positive and negative bus bars, both the positive and negative battery parallel cables, or is just dis-connecting the positive battery cables alright? (when discharging/ charging each of the three string separately)

The cost really depends on where the cells are sourced, the BMS used and the capacity required. For the bank I constructed using 32 × 100 Ah prysmatic cells, 4P 8S arrangement the cell cost directly from the manufacturer was $2080, shipping  $250, BMS $100, second passive BMS for its display  $110 and a 3.65V single cell charger  $65 for a total of ~ $2600  for a 10 Kwh bank or $260 perror Kwh nominal . This is about double the cost locally for LA 400Ahusing golf cart batteries, but the comparison is skewered because LA only has <50% useful capacity to get somewhat decent cycle life, as opposed to 70% conservative with LFP.

The Discover AES is $7200 for 6.6Kwh or ~$1000 per Kwh nominal, but does have a 10 year warranty, communication ability with certain equipment wrapped in a nice container, should these things be important. The cell cost in the US for reputable cells is about $125 for a 100Ah cell so a DIY bank would be ~$500 per Kwh. https://www.ev-power.eu/CALB-40Ah-400Ah/CALB-CA100FI-Lithium-Cell-LiFePO4-3-2V-100Ah.html In the end it comes down to personal preference and requirements. The Discover AES claims 6.6Kwh usable which suggests it too may be closer to 10Kwh actual, if so the price would be more attractive.

This link has some useful information.  
http://nordkyndesign.com/assembling-a-lithium-iron-phosphate-marine-house-bank/

Estragon

I'm also considering LFP for when (hopefully in the far distant future) I need to replace FLA.  In my climate, low temp charging is an issue though, so I'm thinking I might start with small "project" banks at first to get the charging and control stuff figured out while limiting the cost of mistakes.

A security light project might be a place to start