diy solar

diy solar

Integrating DIY LFPs with FLA and Adding Solar

Thom Price

New Member
Joined
Apr 9, 2020
Messages
35
I moved aboard my boat three months ago in Baltimore Harbor, and I liveaboard full-time. I absolutely love it. I've made several upgrades but have a long way to go to get things the way I want them. Energy wise, I've made a huge difference by replacing the 36 year old 21 cubic foot refrigerator with a new one and replacing all of the lights with LEDs. Now I'm ready to go solar.

What I have:
  • 1984 Hatteras 53' Motor Yacht
  • 32 V DC used for starting engines (2 Detroit Diesel 8V-71TI), running hydraulics, pumping water, flushing toilets, and some lights
  • Two 32 V, 1000 AH FLA battery banks made up of four 8 V batteries each
  • An Outback 3.2 KW inverter/charger
  • A 6 KW diesel generator for AC power
  • 50 A shore power service
  • A brand new Energy Star refrigerator and dishwasher, induction cooktop, microwave/convection (1.5 KW), and a small pizza oven (1.5 KW), which are all 120 V
  • A newish stacked washer/dryer (120 V)
  • Four reverse cycle air conditioner/heaters (240 V)
  • A 20 gallon, 240 V water heater (kept turned off most of the time because the marina has nice showering facilities and the dishwasher heats its own water)
  • A flat screen TV (Energy Star) and various small electronics
  • A forced air diesel fired heater (17 KBTU) and several small portable electric space heaters
  • Two 100 W monocrystalline panels that charge a 50 AH LFP battery for my electric dinghy
  • An electric bike used for daily commuting
What I want to achieve:
  • Reduce my need for shore power to 150 KWH/month on average when in the marina (currently 300 KWH baseload plus heating and cooling)
  • Use the generator no more than 2 hours/week when at anchor (presuming nice weather)
  • Find the most cost effective solution over a 10 year span
How I am hoping to do it:
  • Remove one FLA battery bank and replace it with LFPs
  • Add 2.4 KW of solar panels above the fly bridge
  • Insulate between head liners and outside
Rationale:
  • Batteries - Three of the FLA batteries need to be replaced and they cost around $600 each. It would be better to replace 4 of them so that would be $2400 if I stayed with FLA. I need one bank of FLAs to crank the engines and raise the anchor, but the other bank can be used for everything else. Since I'm wanting to use as much solar as I can, I'll be cycling the batteries frequently and so LFPs should be more cost effective over a 10 year span, especially since I'm getting 3.2 V prismatic cells from China and spending $1040 for a 280 AH bank to start off with. I may double that if needed, but I figured it would be better to start small and grow since prices tend to fall for these things. The batteries are arriving next week so now I have to get the rest of the system figured out (hence this post).
  • Solar Panels - I have a space in front of the radar arch that's 10' x 12', which currently has a canvas Bimini on tubular frames that I can cover with panels. The canvas is getting old and I'm thinking that the most cost effective thing would be to replace it with rigid panels fastened to longitudinal rails added to the existing structure. I can add more panels behind the radar arch (1.8 KW), but again want to start smallish and grow as needed.
  • Insulation - This is just part of improving the energy efficiency of the boat by reducing the heat transfer and hence the need to heat and cool as much. I'll be removing the head liners to reveal 1.5" cavities. I'll take a sheet of 1/2" foil faced polyisocyanurate foam board and glue it to the roof, then leave a 1" air gap below that, then install another 1/2" layer of foil faced foam board below the beams and install a new ceiling surface below that. In the galley it will be "stainless steel" contact paper which should reflect the radiant heat back down. I think I'll go with copper in the master state room and white bead board in the salon.
What I need your help with:
  • BMS - I think I need a 10S BMS for the LFP cells. Do I? What do I need to look for?
  • Solar Charge Controller - I was thinking of going with Outback, since I already have an Outback inverter/charger (click for details) for the FLA bank. I want to keep the two banks separated as much as possible, but it would be good to not have a hodge podge of brands. Then I guess I would need to get their Hub to link things together and the Mate, which is where you set all of the parameters. Does anyone here have experience with this brand? Which parts would I need to get? Should I just go with another brand and let the Outback take care of the FLAs?
  • Second inverter/charger? - I guess it would be better to have a dedicated inverter/charger for each bank since the chemistry is different and float charges different. I could have used 11 cells in my LFP bank and gotten away with one inverter/charger for both chemistries but I realized that after ordering the batteries. I could also go ahead and order 12 more cells from China for a total of 22, and build a 11S2P 560AH bank - spending about the same as I would on a second charger/inverter. I'll probably need them. Should I just do that or get a second inverter/charger?
  • Paralleling the two banks - Though I want to keep the two banks separate for the most part, there may be times that it will be useful to parallel them, like when I want to top off the FLAs from the LFPs when at anchor. I have a button at the helm to do that temporarily when starting if needed, but it only connects them while the button is being depressed. Would this be too hard on the LFPs to parallel them with the FLAs when starting the engines? What hardware would I install to top off the FLAs from the LFPs?
  • What's the best way to maximize the solar and minimize the shore power? I'm thinking there must be a controller which would turn off the shore power when certain conditions are met like batteries being full and plenty of sunshine, then turn it back on when the batteries are low. I'd like to start off a sunny day with the batteries depleted, ideally. Other than doing all of that manually, is there a better solution?
  • What am I not asking that I should be asking? Is there some component I have overlooked?
Thanks!
 
Yes. The BMS obviously has to support the right number of cells.

I'm just surprised that you would consider something where we already discussed the mismatch in operating voltages between LFP and FLA/AGM in a 32V system.

It's basically a bad idea all around.
 
Yes. The BMS obviously has to support the right number of cells.

I'm just surprised that you would consider something where we already discussed the mismatch in operating voltages between LFP and FLA/AGM in a 32V system.

It's basically a bad idea all around.
Operating voltage or float voltage?
So, we discussed the FLA needing a higher float voltage than the LFP can handle, and that if the two batteries had different chargers that wouldn't be a problem, or if I were to use 11 cells for the LFP it wouldn't be a problem. Those are the two scenarios that I am evaluating here.
Does that make more sense?
 
Unless your voltages between the FLA bank and LFP are "REAL" close, you can expect a massive inrush when connecting the banks. just FYI.
 
If you're having trouble finding DC switches for the LFP to FLA bank, this may help:

https://www.industrialcontroldirect...onnects-317/?zenid=hh6kps6c7m8dbq6btoijd09a56

I had a lot of difficulty finding a 3 position, 4 pull DC Switch rated for 30 amps and 45 volts. I eventually found some AC rated switches, but that does not mean DC. I was given that website as a place to go for higher amperage / voltage DC switches. I had tried Amazon, Ebay, and google, and not found what I was looking for. On that website, I found one, but the price was $145, and that was more than I wanted to pay for the project I was doing.
 
Operating voltage or float voltage?
So, we discussed the FLA needing a higher float voltage than the LFP can handle, and that if the two batteries had different chargers that wouldn't be a problem, or if I were to use 11 cells for the LFP it wouldn't be a problem. Those are the two scenarios that I am evaluating here.
Does that make more sense?

yes.

Unless your voltages between the FLA bank and LFP are "REAL" close, you can expect a massive inrush when connecting the banks. just FYI.

This.
 
OK, here's the direction I'm going in:
  • FLA bank is charged by the alternators and used to start the engines and raise the anchor, but not connected to the house bank. No paralleling.
  • The shore power and generator is connected to the existing OutBack charger/inverter which will in turn be connected to the LFP bank.
  • The LFP bank will be comprised of 10 cells connected to a BMS (that still needs to be discovered).
  • The LFPs will be charged by an Outback FlexMax 80 solar charge controller which, along with the inverter is connected to an OutBack Mate2 user interface, and an OutBack FlexNet DC systems monitor.
  • The solar charge controller will be connected to 8 - 72 cell monocrystalline panels arranged in 2S4P.
This system will certainly work, and I can easily add 10 more LFP cells as needed, or even more panels without having to add another solar charge controller.

My questions are:
  • Can I find/make a 32V DC-DC charger that I can use to top up the FLA bank from the LFP bank, so that there isn't "a massive inrush when connecting the banks"? That way the FLA bank doesn't risk sitting at a lower state of charge if I'm not running the engines for long periods of time.
  • Is the amp rating of a BMS related more to the discharge capacity of the batteries or to the maximum amp draw they will likely need to provide? These cells can provide 280 amps, but I hope I never need more than 60.
 
I think a 32 volt DC-DC charger will be a hard order to fill, but if I do a 12 volt battery bank to 24 volt battery bank, I plan on using something like a Victron DC-DC charger to charge a 12 volt breakaway battery on my RV. https://www.victronenergy.com/blog/2019/10/10/new-product-orion-tr-smart-dc-dc-charger/

THere’s some good info there, including a link to a blog that talks about how it protects an alternator.

I had a bit of trouble locating this charger, and 12 volt and 24 volt are pretty standard batteries, but I noticed it got hard to locate anything other than those two voltages when I searched, I was thinking of going to 36 volt battery bank, but that’s such an odd size, I didn‘t find anything I wanted, and there were some 48 volt parts if I upgraded to that, but I don’t think I need that much power for my Boondocking RV. I want 48 volts to run my Duel AC for my 35’ RV, but I just can’t fit the panels needed to do it.
 
You likely need to look for an 8-16S BMS. They'll work with any number of cells in that range. You're likely well outside the <$100 field of options and looking towards the higher end options.

If your FLA only start the engines, then they should be fine for a week or two after a brief alternator charge.

32V is just odd, but something like this might work:


This has a huge DC input range (12-90V) and a programmable output range. Power it with the LFP bank directly and set it to your FLA float voltage (about 36V?) to keep the FLA battery floated.

BMS amps relate to the actual current passing through the BMS. It depends on the system.

If you have a 2000W inverter with 32V battery, 2000/32 = 62.5A choose something notably higher. BMS often have different ratings for charge and discharge. Make sure they both work for you.
 
What voltage difference are we talking? Its not hard to calculate/estimate the peak cross over current when connecting the two batteries together. Take the resting voltage of both batteries, then add the internal resistance of each battery in series, then calculate the resistance of the wiring between. That will yield a voltage differential and a bunch of series resistors to sum up. If necessary you can reduce the wire gauge between the two banks to limit current.

Remember that lead banks have the peukert effect and higher internal resistance. A deeply discharged lead bank will actually accept less current than a 50% DOD one, and the current peak will only last a few seconds as a surface charge builds.

The inverse, connecting a discharged LFP bank to a charged lead bank also will be limited by the lead banks internal resistance. Its a simple matter to use a load to find the banks IR, then find the voltage drop. For example 280AH cells will accept about 100A at 5-10% SOC, with a terminal voltage of around 3-3.05V.

Its totally possible to limit cross current with wiring gauge. You just need to do a little math, and possibly some testing on your lead bank.

As a note "In-Rush" typically refers to momentary peak current (less than 1 second), when switching a load/source. This is mostly for calculating switch limits, etc. Fuses, breakers, and wire don't care about such short time periods. Instead you are looking for the sustained current, typically over 5-20 seconds for fuses, and 10 minutes for wire.
 
If you are just looking to equalize the two banks, you can also switch a big resistor to limit current between the banks (same as the wire concept). This can be a coil of modest gauge wire in a bucket of water for example. I don't think this will be necessary if the banks have nominal voltages within 1V of each other. However, if you go with a lower cell count than typical, 2V or greater nominal voltage difference could result in significant currents with a discharged LFP bank.

If you just want to keep your lead bank topped up, then get a DC-DC power supply, and adjust it to a good float voltage for the lead bank. There are many flexible options.

Another option, if your LFP bank is at a higher voltage, is to use a solar PWM controller, and connect the LFP side to the panel input. Gen-A-Sun also makes a MPPT which actually can boost to a higher voltage than the input, which would work even if the LFP bank is a lower voltage.
 
When I measure a healthy 12V battery, I get a measurement of about 3-5mΩ. Yes, FLA/AGM batteries have higher resistance than LFP, but it's still a very, very low resistance.
 
When I measure a healthy 12V battery, I get a measurement of about 3-5mΩ. Yes, FLA/AGM batteries have higher resistance than LFP, but it's still a very, very low resistance
I’m curious how you measure resistance on a battery? Probably not an ohmmeter, I tried measuring ohms many years ago on a live switch, and what I created was an arc welder. Melted the tip of my probe off.? That was three phase power, I can’t remember if it was 120 VAC, 240 VAC,or greater.
 
You can good the process, but essentially you put a known load on the bank and measure the voltage drop, than put another small known load. The delta V and Delta I can be used to find R, V=I*R
 
Or you can buy devices specifically designed to measure the internal resistances of batteries.
 
Those IR meters are good for comparison purposes, but they won't give you the true DC IR, which is ideal for calculating behavior we are discussing. Typically those meters are giving you an AC impedance value, which is going to be a fair bit lower than the DC resistance.
 
I'm speaking comparative as well. I've measured exactly 420 Lithium NMC cells and 63 LFP cells. Lithium tends to be a little lower.
 
You likely need to look for an 8-16S BMS. They'll work with any number of cells in that range. You're likely well outside the <$100 field of options and looking towards the higher end options.

If your FLA only start the engines, then they should be fine for a week or two after a brief alternator charge.

32V is just odd, but something like this might work:


This has a huge DC input range (12-90V) and a programmable output range. Power it with the LFP bank directly and set it to your FLA float voltage (about 36V?) to keep the FLA battery floated.
Thanks for this link! I've already thought of several things I can use this for.
 
Back
Top