270ah DIY LiFePO4 build

[vas]

So I will probably use the relay on my Victron BMV712 to sound an alarm to remind me to end charging. It's not critical if I forget as the charging sources will stop around 90-95% (I need to experiment) and of course, the final backstop is the high voltage disconnect of the BMS but I don't want to rely on that.

For the lower limit I think I'll alarm on 20% and again there is the low voltage disconnect as a backstop.

All this is probably going to be clearer when I've done some sea trials.

Oops!
That's going to be a definite show stopper for many, me included!
Fe, right now it's peak summer and in in a non boaty holiday for 10days. Boat is left idle, with one fridge running plus BMS (as in building management system, not battery!) alarm and SMS messaging for emergencies. These things don't consume much.
How would I setup an unattended charging regime from my 600w solar?

Iow, is there an easy and failsafe way of running some loads unattended?
If micros and arduino could be of help I'd be happy to contribute. Does nt look too simple from the voltage pov, however if accurate temp monitoring (ambient vs internal) is helpful a system could be V.easily implemented!

V.

 

[Poey50]

Oops!
That's going to be a definite show stopper for many, me included!
Fe, right now it's peak summer and in in a non boaty holiday for 10days. Boat is left idle, with one fridge running plus BMS alarm and SMS messaging for emergencies. These things don't consume much. How would I setup a charging regime from my 600w solar?

Iow, is there an easy and failsafe way of running some loads unattended?
If micros and arduino could be of help I'd be happy to contribute. Does nt like too simple from the voltage pov, however if accurate temp monitoring (ambient vs internal) is helpful a system could be V.easily implemented!

V.

You should be able to set a good configurable solar controller like a Victron Smart Solar to manage that but you'd need to experiment with the voltage. Something around the 13.3v or 13.4v mark will keep them charged but not full. People leave their boats in this way successfully once they have worked out a method. I sound a little vague as I haven't yet done it myself.

Charging at 13.2v is said to keep about 40% SOC by some (ideal for LFP) so there probably is a sweet spot to be found for settings when leaving the boat depending on loads.

Temperature is a non-issue at the low charge and discharge rates when leaving a boat. It's where boats - at least those afloat - have an advantage over RVs used in hot desert conditions.

 

[MarkCX]

I’m a little confused. If one of the advantages of lithium is the ability to use a significantly greater amount of a battery’s Ah, why would you want to keep it it 40% SOC? That would give the same useable capacity as lead acid. Or am I missing something?

 

[Poey50]

I’m a little confused. If one of the advantages of lithium is the ability to use a significantly greater amount of a battery’s Ah, why would you want to keep it it 40% SOC? That would give the same useable capacity as lead acid. Or am I missing something?

Only for long term storage or maintenance of a few loads when leaving the boat. While the happy place for a resting lead acid is 100% SOC, LFP rests with least deterioration in the middle of the range. But I don't mean to over-emphasise that - the main thing is not to let LFP hang around needlessly at full charge. As there is more available capacity and no voltage sag at lower states of charge (as with lead acid) then people sometimes just leave a controller on the lower storage setting if going for a day sail or weekend trip and only increase the end of bulk voltage setting if off for a longer cruise. Rod Collins describes testing the ability of his LFP pack to start his engine at a very low state of charge. It did so without difficulty repeatedly until he got bored with doing it.

Day to day usage is quite different to lead acid. With lead acid we are mindful (or should be) of the batteries returning to full charge at the end of a trip to give them a chance of a full and healthy life. With LFP the opposite is true. A couple of years ago I saw someone selling a relatively new drop-in LFP second-hand and I enquired what state of charge it was at. "Don't worry' came the reply "I've kept it fully charged". Not everyone knows the difference.

But, as previously said, I wouldn't hesitate to charge to 100% and use the full range occasionally as needed. But I don't need that full range very often and it is good to have it as a reserve.

 

[MarkCX]

Ok. That makes more sense. Some way of remotely allowing it to get nearer to 100% prior to your arrival at the boat might be useful.

 

[Poey50]

Ok. That makes more sense. Some way of remotely allowing it to get nearer to 100% prior to your arrival at the boat might be useful.

Yes, that would be cool!

But for me, if I leave the boat at 60% state of charge and no loads and no charge then given a 3% per month rate of spontaneous discharge I could go back in three months time and still have more usable capacity than I would now with my two lead acids maintained at 100%. And that's not including the ability of the LFP to take charge far more quickly than lead acid in its upper 20%. From my remaining 55% to 100% would be quick. With a Balmar alternator and external regulator I could put 100 ah back in an hour of motoring. By the time I left Chichester Harbour if motoring I might be around 85%.

 

[gregcope]

Oops!
That's going to be a definite show stopper for many, me included!
Fe, right now it's peak summer and in in a non boaty holiday for 10days. Boat is left idle, with one fridge running plus BMS (as in building management system, not battery!) alarm and SMS messaging for emergencies. These things don't consume much.
How would I setup an unattended charging regime from my 600w solar?

Iow, is there an easy and failsafe way of running some loads unattended?
If micros and arduino could be of help I'd be happy to contribute. Does nt look too simple from the voltage pov, however if accurate temp monitoring (ambient vs internal) is helpful a system could be V.easily implemented!

V.

@vas, If you got a BMS with the ability to control a relay then the BMS could shut off the charge circuit before over voltage disconnect (ie full) becomes an issue. The Smart123 BMS can do this. This is why better LifePo4 designs have separate charge/load circuits so that you can avoid high voltage disconnects and if you do you only disconnect the charge source and not load.

Some of the victron kit has a load circuit that can be enabled under certain conditions.

If you were confident your solar could keep up the fridge could be left on a load circuit and the BMS could either disable charging if it got full or disable the load if the cells approached low voltage.

@Poey50 Has that BMS and relays in his design.

If your looking for some sort of more dynamic "Keep it at 40% when unattended and higher when on baord" thats a bit more complex. I know the Smart123 has a Serial console that some plug RPIs into, not sure if you can change its settings from there.

 

[Poey50]

@vas, If you got a BMS with the ability to control a relay then the BMS could shut off the charge circuit before over voltage disconnect (ie full) becomes an issue. The Smart123 BMS can do this. This is why better LifePo4 designs have separate charge/load circuits so that you can avoid high voltage disconnects and if you do you only disconnect the charge source and not load.

Some of the victron kit has a load circuit that can be enabled under certain conditions.

If you were confident your solar could keep up the fridge could be left on a load circuit and the BMS could either disable charging if it got full or disable the load if the cells approached low voltage.

@Poey50 Has that BMS and relays in his design.

If your looking for some sort of more dynamic "Keep it at 40% when unattended and higher when on baord" thats a bit more complex. I know the Smart123 has a Serial console that some plug RPIs into, not sure if you can change its settings from there.

You can certainly use the BMS in that way and 123Electric sort of encourage that use in their description of their 'normal' setting for high voltage disconnect. But I think there is a good argument on boats, where the risks are higher, to only use the BMS as a last line of defence to provide catastrophic protection - a bit like a fuse in an electrical system. This is certainly the message from Nordkyn and from Marine How To, that one's control systems for charging should be separate from the final protection systems. But I'm someone who finds the use of belt and braces to be recklessly unprepared. You can probably make any system work if you understand it.

 

[Poey50]

As we are beginning to talk about charging systems this seems a timely point to return to Nordkyn Design's 5th principle for the design of an LFP marine system, my DIY pack as described covering the first four (see the original post).

5. All charging sources and regulators that will ever feed the new battery need to be re-assessed for suitability: either they can be reconfigured to operate acceptably, or they will need to be replaced.

This is a simplified diagram.

Last year I bought one of the Sterling 60amp Battery to Battery Chargers, on clearance with a short warranty, for a good price. The original plan was to charge the Red Flash AGM start battery from my standard Nanni 100ah alternator and then charge the LFP from the start battery via the B2B using LFP specific settings. The intention was to use that for a season and then consider a big upgrade to charge the LFP direct from a Balmar alternator and external regulator and use the B2B to charge the start battery from the LFP. The original plan is part of the aforementioned £1500 project cost. The recommended headroom for an alternator used with a B2B is for a rating 40% higher than the B2B so this should have worked but at the cost of not using one of the big advantages of LFP which is its ability to take plenty of charge direct from the alternator. 100amps of charge is only just above 0.3C for 270ah of LFP so that is still quite a moderate rate of charge without undue concerns about heat.

This all changed two weeks ago when the Nanni alternator started sending 15.4 volts to the start battery at tick-over and it proved impossible to find a replacement for the 14 year old Slovenian alternator regulator. So that brought the upgrade forwards by a year and, added £800 to the overall cost for the 100amp Balmar + MC614 external regulator + temperature sensors and another £450 (!) for a serpentine belt upgrade kit which I had to buy from the US. Balmar advise that a V belt is fine for their alternators up to 100amps but I had already seen too much belt dust from previous boost charge systems working the alternator hard and, anticipating how relentlessly the low resistance LFP would work the alternator, a serpentine belt seemed wise, though expensive.

 

[Poey50]

To complete the charging section, the three charging controllers - Balmar MC614, Victron Smart Solar, and Victron BlueSmart Mains Charger - all have advanced settings allowing user-configuration. All three will be similarly configured to charge up to the end of bulk at 13.8volts, to have no absorption period (or as short an absorption as possible), no temperature compensation, and a float of 13.2volt which effectively is no float at all. None of the chargers will be controlled by the BMS but are autonomously controlled through their own settings. As previously said I will set the Victron BMV712 relay to trigger an alarm if these settings are exceeded (through a controller fault) and the BMS is set to disconnect the charge relay should I fail to respond to the alarm.

There will be two additional protections in relation to the alternator charging, which given its power can do more damage, more quickly. With the engine running I can't guarantee to hear an alarm and a V-Guard II will trigger a relay to interrupt the ignition feed to the Balmar regulator at 14.00 volts.

V-Guard II – programmable battery monitor 12V and 24V, with LCD: Amazon.co.uk: Car & Motorbike

The final line of protection from the alternator is again the BMS high voltage disconnect which will end alternator (or any other) charging if any cell exceeds 3.65 volts. To protect the alternator from the resultant voltage spike i am fitting a Sterling Alternator Protect.

Sterling Power Alternator Protection Device - APD24 | eBay

And that's it. I will need to trial those voltages to get the state of charge needed. One risk is that the LFP will not be charged high enough for the passive balancing of the BMS to work. But if cells get out of balance then I can reconfigure the mains charger to occasionally do a higher charge or the active balancer I added to the build will take over if the cells vary by more than 0.1volt But having done a careful top-balance during the build, I think them going out of balance quickly is unlikely.

Now to fit it all together.

Thanks for staying the course, if you have. Comments and questions welcome.

 

[pcatterall]

great thread! The sort that is great to use as a reference for those considering moving to lithium.
With my particular battery needs it would be a waste of time and money ditching my lead acids,. I could either save weight and space and retain the same capacity or massively increase my capacity for the same weight/space. But, in my case, the space/weight gain is not important and I couldn't use the extra capacity! Added to this is the complexity of the charging/discharging regime and of course cost.
So not for me on the boat but possible use in other projects.
I appreciate that the thread purpose was not to debate the pros and cons so apologies in advance!

 

[vas]

OK, looks good.

the isolation of the charging vs load circuits in the BMS would be a bit more complicated if you had a combined mains charger/inverter like the Multiplus 3000/24 I have though...
You'd need a set of serious relays so that you can isolate the charging circuit and only run the inverter, or somehow run both which I really cannot get my head round it atm.

There's only one set of M8 (iirc) posts where you connect the batteries and is used both for pumping Amps to them or drawing from them. Most important it can do a bit of both, like when you have a big starting load whilst running the generator, it gets all it can from the generator and covers any peaks from the batteries.
That, I don't think will be possible!

V.

 

[Poey50]

OK, looks good.

the isolation of the charging vs load circuits in the BMS would be a bit more complicated if you had a combined mains charger/inverter like the Multiplus 3000/24 I have though...
You'd need a set of serious relays so that you can isolate the charging circuit and only run the inverter, or somehow run both which I really cannot get my head round it atm.

There's only one set of M8 (iirc) posts where you connect the batteries and is used both for pumping Amps to them or drawing from them. Most important it can do a bit of both, like when you have a big starting load whilst running the generator, it gets all it can from the generator and covers any peaks from the batteries.
That, I don't think will be possible!

V.

Yes, an inverter charger doesn't sound like an immediate fit with the principle of separating charge and load. But they are popular so there will be some work-arounds. It's not something I've explored in any detail.

 

[Poey50]

Here's the Nordkyn Design alternative suggestions when load and charge can't be separated. Both use a hybrid system making use of a lead acid left in the system on disconnect.

Full details at http://nordkyndesign.com/electrical-design-for-a-marine-lithium-battery-bank/

 

[gregcope]

OK, looks good.

the isolation of the charging vs load circuits in the BMS would be a bit more complicated if you had a combined mains charger/inverter like the Multiplus 3000/24 I have though...
You'd need a set of serious relays so that you can isolate the charging circuit and only run the inverter, or somehow run both which I really cannot get my head round it atm.

There's only one set of M8 (iirc) posts where you connect the batteries and is used both for pumping Amps to them or drawing from them. Most important it can do a bit of both, like when you have a big starting load whilst running the generator, it gets all it can from the generator and covers any peaks from the batteries.
That, I don't think will be possible!

V.

For larger loads/charging I have seen multiple batteries used to cover those larger loads. Some people with induction hobs have three or more batteries. I think Sailing Uma and Delos are like this. They may install smaller capacity and parallel up. Obviously cells do not need to be 280ah ones, but this would require multiple BMS that can be a significant part of the costs.

 

[Zing]

You should be able to set a good configurable solar controller like a Victron Smart Solar to manage that but you'd need to experiment with the voltage. Something around the 13.3v or 13.4v mark will keep them charged but not full. People leave their boats in this way successfully once they have worked out a method. I sound a little vague as I haven't yet done it myself.

Charging at 13.2v is said to keep about 40% SOC by some (ideal for LFP) so there probably is a sweet spot to be found for settings when leaving the boat depending on loads....

13.2 is 3.3v per cell. That is very close to the resting fully charged level. If you look at a low C discharge curve, a voltage, which corresponds to about 40% is more like 12.8v or 3.2v, or in my experience a bit less as the discharge charts have more discharge shown than usually seen in a stored boat. Small difference in voltage, big difference in capacity. It’s good to store them in a lowish state of charge.

 

[Poey50]

13.2 is 3.3v per cell. That is very close to the resting fully charged level. If you look at a low C discharge curve, a voltage, which corresponds to about 40% is more like 12.8v or 3.2v, or in my experience a bit less as the discharge charts have more discharge shown than usually seen in a stored boat. Small difference in voltage, big difference in capacity. It’s good to store them in a lowish state of charge.

As said I'm quoting others rather than my own experience so I plan to experiment to find the right voltage. 3.3 volts per cell is sometimes said to be a pivotal number. Below that you don't have significant charging, above that you do. The curves are pretty flat so can't be totally relied on for reading off state of charge to voltage. My pack is at 53% SOC at the moment with cells resting at 3.31v.

This is another quote from the notes I've been making over the last year on the experience of others .. "use a charger you can configure as a constant voltage DC power supply and set it for 13.0-13.2V. It will just prevent the LFP battery from going flat. This works best if a small load is always on, so a full battery can gradually discharge after you return to the dock."

 

[Poey50]

To complete the charging section, the three charging controllers - Balmar MC614, Victron Smart Solar, and Victron BlueSmart Mains Charger - all have advanced settings allowing user-configuration. All three will be similarly configured to charge up to the end of bulk at 13.8volts, to have no absorption period (or as short an absorption as possible), no temperature compensation, and a float of 13.2volt which effectively is no float at all. None of the chargers will be controlled by the BMS but are autonomously controlled through their own settings. As previously said I will set the Victron BMV712 relay to trigger an alarm if these settings are exceeded (through a controller fault) and the BMS is set to disconnect the charge relay should I fail to respond to the alarm.

There will be two additional protections in relation to the alternator charging, which given its power can do more damage, more quickly. With the engine running I can't guarantee to hear an alarm and a V-Guard II will trigger a relay to interrupt the ignition feed to the Balmar regulator at 14.00 volts.

V-Guard II – programmable battery monitor 12V and 24V, with LCD: Amazon.co.uk: Car & Motorbike

The final line of protection from the alternator is again the BMS high voltage disconnect which will end alternator (or any other) charging if any cell exceeds 3.65 volts. To protect the alternator from the resultant voltage spike i am fitting a Sterling Alternator Protect.

Sterling Power Alternator Protection Device - APD24 | eBay

And that's it. I will need to trial those voltages to get the state of charge needed. One risk is that the LFP will not be charged high enough for the passive balancing of the BMS to work. But if cells get out of balance then I can reconfigure the mains charger to occasionally do a higher charge or the active balancer I added to the build will take over if the cells vary by more than 0.1volt But having done a careful top-balance during the build, I think them going out of balance quickly is unlikely.

Now to fit it all together.

Thanks for staying the course, if you have. Comments and questions welcome.

Some further exploration on this suggests that setting the end of bulk voltage on the solar controller slightly higher than that of the external regulator will prevent a run of the engine sending the solar controller into float.

 

[GHA]

Another LiFePo4 installation write up here >>

My conversion to Lithium (and Cerbo) - by Stefanie | Victron Energy

 

[Poey50]

That's a lot of blue. I think Stephanie is considerably better endowed in the wonga department than me. Looking forward to part 2 though.