diy solar

diy solar

Trying to put it all together ...

PeterBC

New Member
Joined
Nov 4, 2020
Messages
46
So I’m relatively new to this DIY LiFePO4 batteries but need to get moving as my old 4 x 220AH Victron GEL’s are close to stuffed after only 4 busy seasons on my barge (500 cycles, but too many > 50% I think). I've be scouring these forums for info, and several other sources. I'm an Electrical Engineer but not practicing.
Not quite sure which is best forum to post this be it this one or BMS's etc..

I have a Victron Quattro 5000/24V and 4 x 250W Solar panels with a Victron SmartSolar MPPT 100/50. Charing is also from a big 3 phase 14kw generator which I can charge via the Victron at 2.4kw , 100 Amps into the battery using a single phase, or shore power. Load is primarily AC via the Victron at 5kw max, approx. 220Amps from the batteries. I also have a Victron BMV-7xx battery monitor (and shunt).

I’m planning 24 x 3.2V LiFePO4 280AH cells as they are 1/5th the price or better vrs other retail chemistries or Lithiums for the same useable AH. Better than 1/10th the lifetime cost per kwH cycled than other lead acid chemistries, and generally more tolerant of abuse (as long as a BMS is present and chargers/loads obey any requisite signals).

Key questions:
  1. Which battery ? Xuba, Basen, other ? What price ? What freight to France ? How long to ship by sea vrs Air ?
  2. 8S x 3P or 3P x 8S ? Passive Balance current (35ma or 200ma) of 3P x 280AH -840AH is it enough ? Should Active Balancers be considered (1.5 or 5A) , 1 BMS vrs 3 BMS but can 1BMS handle charge/discharge current of 3P batteries 100/200Amp) ? If 3 banks, size each BMS to handle ½ the charging /discharging load for redundancy.
  3. BMS Daly , Electrodacus , Heltec active balancing , Chargery or Other ?
  4. Low temp cutoff : My Victron(s) have several thermistor methods, I only need to shut off Quattro and MPPT in a low temp case , typically over winter when unattended. I suspect both will need a thermistor to detect low temp state for cutoff.
  5. Failed cells.. If a cell fails can it be replaced with a new cell safely ? Should I buy some spare cells given delivery takes a long time ?
Now as I understand it , normal operation would be that the Quattro and MPPT charger will be set for proper voltages to back off charging when the bank is full , and disconnect if too low. However my BMS must also protect the individual cells (or sets of 3) in case of over or under voltage and cut off any load or charging. They can signal my Quattro and MPPT controller in these cases to stop. Ideally that’s what they do. I’m concerned about a BMS directly across the battery that might cut full charging or load currents could cause considerable spikes.

An Electrodacus I understand can signal my Quattro (and my MPPT at the same time) if needed for overvoltage to shutdown. For low voltage my Victron would shutdown, but I’d also need to disconnect any other DC small loads I may have. I presume the Electrodacus could signal a relay on the DC loads (and the Quattro again) to shutdown if an individual cell was too low.

Is replacing a failed cell with a new one something that’s doable ? If it had substantially better capacity what would likely happen ?

Alternatively I could perhaps use a Daly BMS , probably common port (as my currents are fairly high). If using 3 x 8S arrangement and 3 Daly BMS’s I could configure for approx. ½ the peak load (allowing for one bank to fail and still be able to get by with just 2 banks). So typically 50Amp peak charging and 100Amp peak load in that situation. I’d oversize the Daly to at least be a 75Amp charging and 150Amp peak current I think. 3 Electrodacus's seems a bit $$ and complex ?

Need fuses / circuit breakers, isolator switches for Battery and MPPT and Panels , controlled relays for disconnect of loads and any charging sources that cant be signalled to disconnect (eg MPPT ? only has low temp cutoff, if an individual cell V is too high it needs to be stopped.. I gather via relay on the panel inputs ? ) .. Got get my shopping list sorted including for these bits.

Thanks..
 
If you go 8s3p then you will have three BMS devices. Let's say you went with the Overkill Solar 8s 24v BMS. It supports 100 amp charge/discharge. Three of those in "parallel" gives you 300 amps. Would that satisfy your expected maximum amp draw?

The 8s3p configuration gives you more visibility into the actual cells. 3p8s has the ability to hide a bad cell. It would also require a much higher amp rated BMS since there would be only one BMS in that configuration.

Cell Failure - It isn't common, but if it's something you're concerned about then I would buy at least three spare cells. Should you need to replace a cell then you're going to want to run that battery (all 8 cells) through a Top Balancing process.
 
If you go 8s3p then you will have three BMS devices. Let's say you went with the Overkill Solar 8s 24v BMS. It supports 100 amp charge/discharge. Three of those in "parallel" gives you 300 amps. Would that satisfy your expected maximum amp draw?

The 8s3p configuration gives you more visibility into the actual cells. 3p8s has the ability to hide a bad cell. It would also require a much higher amp rated BMS since there would be only one BMS in that configuration.

Cell Failure - It isn't common, but if it's something you're concerned about then I would buy at least three spare cells. Should you need to replace a cell then you're going to want to run that battery (all 8 cells) through a Top Balancing process.
Thanks for your reply.. Yes I'm leaning to 8s3P with 3 BMS. Yes I think a 100Amp discharge/50Amp charge BMS would be sufficient to allow one of the 3 strings to be out of action and the other 2 could take maximum load. I'd probably increase rating a bit for more headroom. If the BMS disconnects charging directly at the battery due to a cell having high voltage or loads due to a low voltage what happens if thats while full 'load'/charging is underway. Seems like some serious surges could be at play ? If the pack voltage is within limits then the Qattro and MPPT may still be pushing/pulling full boar but the BMC could cut the battery ?
 
Thanks for your reply.. Yes I'm leaning to 8s3P with 3 BMS. Yes I think a 100Amp discharge/50Amp charge BMS would be sufficient to allow one of the 3 strings to be out of action and the other 2 could take maximum load. I'd probably increase rating a bit for more headroom. If the BMS disconnects charging directly at the battery due to a cell having high voltage or loads due to a low voltage what happens if thats while full 'load'/charging is underway. Seems like some serious surges could be at play ? If the pack voltage is within limits then the Qattro and MPPT may still be pushing/pulling full boar but the BMC could cut the battery ?

I don't know for sure, but if the charging is close to the limit, then I would expect the current to be reduced, while the voltage is high. It may be a non-event.
 
Back
Top