270ah DIY LiFePO4 build

[vic008]

A bit OT. But about alternator smart regulators, if they don't mention working with Lithium batteries is it best to assume they are not compatible. Have a Next Step by Amplepower regulator, and it doesn't mention lithium.(haven't gone lithium yet, but will)

 

[halcyon]

A passing question, I do not fancy going through 541 posts, is there any reason for not using a fiber glass moulded box for fitting the battery cells ?

Brian

 

[Poey50]

A passing question, I do not fancy going through 541 posts, is there any reason for not using a fiber glass moulded box for fitting the battery cells ?

Brian

You need to arrange some moderate compression through the pack to constrain swelling during charging. (Unconstrained swelling having been shown to shorten cell life.) So any fixed box may be a problem unless the compression issue is thought about from the start. The usual solution has been rigid end plates compressing via 4 threaded rods. That is how mine is made. The new standard suggested by Rod Collins (who will be a major influence on the ABYC standard, when published, and thereafter on surveyors and insurance) is that the end plates should be non-combustible. Mine are thick ply, which has been common practice. I will change to metal at some point.

 

[Kelpie]

Presumably you could use a slightly oversized box and then have packers to make it a snug fit?
I've been given an old mini safe which is a really nice size, and reasonably fire proof, obviously I'll need to think about sufficient insulation to prevent shorting.

 

[Zing]

A passing question, I do not fancy going through 541 posts, is there any reason for not using a fiber glass moulded box for fitting the battery cells ?

Brian

As mentioned, compressing the cells to prevent their expansion is considered good practice

Flamability is an issue I have always been concerned with. I use steel compression beams on mine and enclose in a melamine lined thick plywood case. Not perfect for flamability. I have the 180 Calb cells. This gives slightly better fire and heat protection than fibreglass it isn't as good as a metal case.

The older style plastic cased Lifepo4 cells like mine use a different chemistry for the electrolyte it seems than the newer aluminium cased ones. Whilst I haven't seen any accounts of that type being shown to have the electrolyte capable of combustion, I have seen it recently for the newer aluminium cased cells (video below). I wonder if that means an different approach to fire protection is necessary.

The aluminium cases also need some extra care in their installation to prevent shorting from the case, also to prevent corrosion of the case.

 

[Poey50]

As mentioned, compressing the cells to prevent their expansion is considered good practice

Flamability is an issue I have always been concerned with. I use steel compression beams on mine and enclose in a melamine lined thick plywood case. Not perfect for flamability. I have the 180 Calb cells. This gives slightly better fire and heat protection than fibreglass it isn't as good as a metal case.

The older style plastic cased Lifepo4 cells like mine use a different chemistry for the electrolyte it seems than the newer aluminium cased ones. Whilst I haven't seen any accounts of that type being shown to have the electrolyte capable of combustion, I have seen it recently for the newer aluminium cased cells (video below). I wonder if that means an different approach to fire protection is necessary.

The aluminium cases also need some extra care in their installation to prevent shorting from the case, also to prevent corrosion of the case.

I'd not seen that so thank you for alerting us. As you say, the need to protect cells from catastrophic damage, that thick smoke and fierce fire all need some careful thought.

 

[Poey50]

Presumably you could use a slightly oversized box and then have packers to make it a snug fit?
I've been given an old mini safe which is a really nice size, and reasonably fire proof, obviously I'll need to think about sufficient insulation to prevent shorting.

As said, as long as you are thinking about compression from the start then several arrangements could work.

 

[Poey50]

A bit OT. But about alternator smart regulators, if they don't mention working with Lithium batteries is it best to assume they are not compatible. Have a Next Step by Amplepower regulator, and it doesn't mention lithium.(haven't gone lithium yet, but will)

It depends on how you plan to use the alternator. If relying mainly on solar then a standard alternator charging the LFP via a B2B is fine as long as the capacity of the alternator is around 50% greater than the B2B. Not generally a problem if using a 30amp Victron Orion which is a popular choice. If you are planning to charge the LFP direct from the alternator then you need a fully user-configurable external regulator like the Balmar MC614 or the Wakespeed 500. I don't even use the recommended lithium profile on the Balmar, preferring a lower target voltage. Most users prefer to choose their own settings since ideas about charging voltages have been revised downwards in recent years.

 

[Kelpie]

After a few delays I got back to the project at last. My cheap JBD BMS seems to be alive! I'm currently charging the pack at 13.8v using my rather dodgy PSU, which is only putting out 2.5A- I was expecting it to go all the way to its rated 10A. If it randomly decides to jump to 20v like it did before, the BMS should now disconnect it.
I had to try a couple of different apps before I found one that worked with my phone and BMS- this aspect of things seems quite glitchy. But it's all running now. Showing my cells at 3.365, 3.360, 3.361, 3.377- for a total difference of 0.017v.
I really should get my finger out and get a replacement PSU, otherwise I'm not going to be able to do a top balance. Need to have the whole system up and running in about six weeks time as we are planning to move aboard the boat. No pressure...

 

[Poey50]

After a few delays I got back to the project at last. My cheap JBD BMS seems to be alive! I'm currently charging the pack at 13.8v using my rather dodgy PSU, which is only putting out 2.5A- I was expecting it to go all the way to its rated 10A. If it randomly decides to jump to 20v like it did before, the BMS should now disconnect it.
I had to try a couple of different apps before I found one that worked with my phone and BMS- this aspect of things seems quite glitchy. But it's all running now. Showing my cells at 3.365, 3.360, 3.361, 3.377- for a total difference of 0.017v.
I really should get my finger out and get a replacement PSU, otherwise I'm not going to be able to do a top balance. Need to have the whole system up and running in about six weeks time as we are planning to move aboard the boat. No pressure...

At that stage of charge the largest difference between the cells isn't very meaningful. Difference only emerges near the top end. Don't be tempted to dispense with the top balance. Plenty do and it stores up trouble.

 

[Poey50]

The American Boat and Yacht Council (ABYC) has now issued technical guidance (TE-13) on LiFePO4 installations. It's only available to members of ABYC and of course it doesn't directly apply to those outside the US. But as they are first out of the blocks then it is likely to be influential to builders, surveyors and, following that, insurers more widely than the States.

The things that have emerged so far are as follows and might be worth bearing in mind if designing a system. Some have already been mentioned.

1. The BMS should communicate a visual and audible warning prior to disconnect.
2. Class T fuses to be used to protect against a dead short.
3. Wooden compression end-boards will not be acceptable. They will need to be metallic because of a fire risk.

Although it isn't apparently listed in the technical guidance it is possible that direct BMS control of chargers may also get into the final standard.

Edit: While these add additional challenges for the DIY builder - especially in terms of selecting a BMS - no drop-in LFP currently on the market will meet #1. This standard may drive changes in BMSs and drop-ins although the boating market is not vast so these might be smaller scale and therefore pricey variants.

 

[Kelpie]

Now that I have a working pack on the bench, I need to translate this in to a permanent installation.
Thinking of using some grp sheet as end plate for compression (I have an old dinghy needing disposed of...)
When it comes to isolators/fuses etc, I'm a bit bewildered by the choices available. What about using a DIN-mount MCB? Would that serve both functions?

 

[Poey50]

Now that I have a working pack on the bench, I need to translate this in to a permanent installation.
Thinking of using some grp sheet as end plate for compression (I have an old dinghy needing disposed of...)
When it comes to isolators/fuses etc, I'm a bit bewildered by the choices available. What about using a DIN-mount MCB? Would that serve both functions?

I'm going to change from 15mm plywood to 6mm aluminium end plates. And I bit the bullet and ordered Class T fuses.

 

[Kelpie]

I've got an old metal safe that is a good fit for the battery pack. I'm weighing up the pros and cons of using it as an enclosure. Obviously it's very robust and fire proof, but it would also reduce ventilation, and there is the risk of a short developing (although the casing is powder coated).

I think I will trust the MCBs for the solar panels and MPPTs, but I will use conventional fuse and isolator on the main power output from the battery.

 

[Poey50]

I've got an old metal safe that is a good fit for the battery pack. I'm weighing up the pros and cons of using it as an enclosure. Obviously it's very robust and fire proof, but it would also reduce ventilation, and there is the risk of a short developing (although the casing is powder coated).

I think I will trust the MCBs for the solar panels and MPPTs, but I will use conventional fuse and isolator on the main power output from the battery.

Despite the well-justified reputation for greater stability of LFP compared to other lithium chemistries, the very low resistance does create the possibility of serious current being delivered into a dead short. I've seen a figure as high as 30C which in the case of my 270ah pack would be 8,000 plus amps. For this reason ABYC recommend Class T for the main fuse. There are concerns about arcing across small gaps of other blown fuse types - apparently or of metal pieces welding together. I'm willing to pay up for the safety margin.

 

[Kelpie]

Despite the well-justified reputation for greater stability of LFP compared to other lithium chemistries, the very low resistance does create the possibility of serious current being delivered into a dead short. I've seen a figure as high as 30C which in the case of my 270ah pack would be 8,000 plus amps. For this reason ABYC recommend Class T for the main fuse. There are concerns about arcing across small gaps of other blown fuse types - apparently or of metal pieces welding together. I'm willing to pay up for the safety margin.

You make a very valid point. Would I be right to think that a risk of a total short only really exists on the main output, or could it occur on the charging side too?

 

[Poey50]

You make a very valid point. Would I be right to think that a risk of a total short only really exists on the main output, or could it occur on the charging side too?

I'm putting mine close to the main positive terminal and the positive divides into charge and load buses downstream of that. At 270 amps mine is still sized to cover the cables.

 

[Poey50]

Good advice from those experienced with DIY LFP batteries is to check that your newly purchased and imported cells are well matched for capacity and internal resistance. Any problems of mismatch could store up difficulties with the pack. Capacity testers are readily available and easy to use but there was no reasonably inexpensive way I knew to measure internal resistance until recently.

This just arrived from Ali Express. 37.99US $ 28% OFF|New High Precision Fast YR1035 Lithium Battery Internal Resistance Test Instrument 100V Electric Vehicle Group 18650|Battery Testers| - AliExpress

Now I've worked out how to change to English from Chinese it is a doddle to use, giving simultaneous read-outs on voltage and internal resistance. It would be easy to check a batch of newly-arrived cells in minutes whereas now to check mine I have to disconnect all bus-bars and other wiring.

Mine came with a Chinese manual only but an English manual is downloadable from this page together with an unboxing video. Yaorea YR1035+ Battery Internal Resistance Meter Tester, Manual

It was one of the Off-Grid Garage videos that alerted me (below). Andy - is a bit confused by the findings but there is a lot of useful information in the comments and, clearly, proper comparative measurements need to be done without busbars.

This may all seem a bit OTT and nerdy but the grey Chinese import market is a jungle and you can get ripped-off. £80 worth of test equipment (capacity and internal resistance) is, I think money well spent to ensure you get what you paid for. And if you need to check on the state of health of your cells in a few years time you will be well-equipped.

 

[Poey50]

I'm in two minds whether to resurrect this thread. On the one hand it is very long, on the other it holds a lot of information in one place.

Anyway I think this LFP topic is relevant to this thread as I seem to have spent quite a lot of time issuing dire warnings about drop-in LFP batteries. I won't repeat myself (much) but these are, at the same time, the most tempting solutions for boaters and the most problematic. As said several times, the safety of a system can be tested with a thought experiment - if you are crossing a busy shipping lane at night, how will you know your LFP pack is about to disconnect and isolate itself and what will happen to your system when it does?

A step towards making drop-in LFP safer on boats has been taken by Rod Collins of Marine HowTo. One big problem with drop-ins is that during alternator charging a BMS disconnect dumps the alternator load and can take out the alternator diodes and, because the load and charge buses are not separated as recommended, the huge voltage spike can also take out marine electronics. Until now the main recommended solution was for LFP drop-ins to be charged via the engine battery with a battery to battery charger. But Collins has a simpler and cheaper solution to the B2B, one that allows drop-ins to be charged direct by an alternator with external regulator. It uses a solenoid connecting house LFP to lead acid starter. It switches on with the ignition circuit and always retains a lead acid starter battery in the system to absorb spikes, keeps the start battery in good health, and keeps essential loads running from the start battery in the event of a disconnect. This setup still doesn't of course provide warnings of disconnect (although that could be built into the system) but it does open up further options for drop-ins to be charged more safely at higher rates. The advert is below and it come with interesting blurb and a downloadable manual. Click on the downward pointed arrow at the end of the blurb.

CMI LiFePO4 150A Load-Dump & Stay-Aive Protection Solenoid

 

[TernVI]

I'm in two minds whether to resurrect this thread. On the one hand it is very long, on the other it holds a lot of information in one place.

Anyway I think this LFP topic is relevant to this thread as I seem to have spent quite a lot of time issuing dire warnings about drop-in LFP batteries. I won't repeat myself (much) but these are, at the same time, the most tempting solutions for boaters and the most problematic. As said several times, the safety of a system can be tested with a thought experiment - if you are crossing a busy shipping lane at night, how will you know your LFP pack is about to disconnect and isolate itself and what will happen to your system when it does?

A step towards making drop-in LFP safer on boats has been taken by Rod Collins of Marine HowTo. One big problem with drop-ins is that during alternator charging a BMS disconnect dumps the alternator load and can take out the alternator diodes and, because the load and charge buses are not separated as recommended, the huge voltage spike can also take out marine electronics. Until now the main recommended solution was for LFP drop-ins to be charged via the engine battery with a battery to battery charger. But Collins has a simpler and cheaper solution to the B2B, one that allows drop-ins to be charged direct by an alternator with external regulator. It uses a solenoid connecting house LFP to lead acid starter. It switches on with the ignition circuit and always retains a lead acid starter battery in the system to absorb spikes, keeps the start battery in good health, and keeps essential loads running from the start battery in the event of a disconnect. This setup still doesn't of course provide warnings of disconnect (although that could be built into the system) but it does open up further options for drop-ins to be charged more safely at higher rates. The advert is below and it come with interesting blurb and a downloadable manual. Click on the downward pointed arrow at the end of the blurb.

CMI LiFePO4 150A Load-Dump & Stay-Aive Protection Solenoid

You seem to be advocating using the lead acid start battery to protect the system by absorbing spikes.
The page you link to really doesn't say that, there is a Sterling alternator protector in the diagram to do that.

Anyone who's looked at desulfator circuits with an oscilloscope will know that lead acid batteries are not really that great for absorbing spikes.

What keeps the vessel alive if the 'drop ins' disconnect when the engine isn't running?

There is a lot of half baked stuff in this.