Wiring for Lifepo4

Ceirwan said:
Where is the 20,000 figure coming from? Not disputing it, but it seems quite generalised.
Presumably it has to depend a lot on cell capacity, if there are multiple cells wired in parallel, etc?


For example a mini 50amp hour LifeP04 battery is going to put out a lot less current during a direct short then a 800ah battery.
Click to expand...
I read a good article a while back that suggested that once you hit 280Ah of lithium battery, AIC becomes an issue. You won't have a problem with a 100Ah lithium battery on a mega fuse.
I run a pair of 280Ah batteries at 24v (so 560Ah each in 12v money). Both fitted with isolators and Nh fuses. It's not worth taking a risk when those fuses are good practise anyway.
My 45 year old boat was fitted with NH fuses for domestic and engine batteries from new. It did have 1000Ah (@12v) worth of lead acid back in those days
 
geem said:
I read a good article a while back that suggested that once you hit 280Ah of lithium battery, AIC becomes an issue. You won't have a problem with a 100Ah lithium battery on a mega fuse.
I run a pair of 280Ah batteries at 24v (so 560Ah each in 12v money). Both fitted with isolators and Nh fuses. It's not worth taking a risk when those fuses are good practise anyway.
My 45 year old boat was fitted with NH fuses for domestic and engine batteries from new. It did have 1000Ah (@12v) worth of lead acid back in those days
Click to expand...
It make sense, and the NH fuses are actually priced fairly reasonably as well.

On that note, can anyone recommend a reasonably priced fuse base for an NH1 class fuse? I've found a few, but ideally I'm looking for one that comes with a little integrated flip off cover or something like that. I found one, but the price is hundreds of pounds.
Failing that I can probably make my own cover if I had to.
 
Ah yes, the prospective short circuit current (PSC) debate… A lot of threads (also on other forums) keep getting stuck on this point, so I thought I’d take a look at the academic evidence on the subject; here is what I discovered: Cell datasheets only ever publish a 1kHz impedance measurement for internal resistance (IR), but unfortunately, this figure is not the same as a cell’s DC IR. A 1kHz impedance measurement is useful for predicting a cell’s performance when driving high frequency inverter-drives (as in an electric car), and it’s also an easy measurement to make for rapid pass / fail quality checks; and that is why it’s in common use.

To estimate PSC, you need a cell’s DC IR, and from my investigations, it seems that this is typically about 25% greater than its1kHz IR measurement. So, it’s possible to estimate PRC from a cell’s datasheet as follows: A cell with a 1kHz IR of 300 micro-ohms will have a DC IR of about 375 micro-ohms. In order to use a fuse, there needs to be a minimal circuit of a cable plus two joints (one to the cell, and one to the fuse). Taking the cable resistance as zero ohms, and a minimum achievable joint resistance of, say, 50 micro-ohms gives a total circuit resistance of 475 micro-ohms. The cell PSC can be then be estimated as 3.2 V / 475 micro-ohms, or 6,736 A.

Putting cells in series doesn’t change PSC, but putting two cells in parallel doubles it to 13,472A (and four in parallel to 26,944A etc.); that is why high AIC fuses such as Class-T or NH are required by ABYC and ISO. A battery’s PSC depends a lot on how many cells are connected in parallel, so a 10kA AIC MRBF fuse might be ok for 1P4S batteries (just about), but ANL or Mega fuses are not suitable. Interestingly, 1,000Ah Winston Thundersky cells have a datasheet IR of 300 micro-ohms, so in theory a 12V (12.8kWh !) battery built with these cells could be fused with an MBRF fuse. In other words, PSC is not necessarily directly related to Ah capacity.

In fact, one of the reasons for building LFP banks with a small number of large cells, rather than paralleling multiple batteries (see Rod Collins and others), is to limit PSC, making this arrangement arguably safer. Other benefits include: simplicity, less need for cell balancing, fewer busbars and joints, less cabling, fewer BMSs and fuses, greater reliability and a more compact construction etc. I gather that EVE 600Ah cells will soon be available, so a 12V battery built with just four of these cells (plus a BMS of course) will probably be all the house battery that most UK boaters will ever need.
 
migs said:
Ah yes, the prospective short circuit current (PSC) debate… A lot of threads (also on other forums) keep getting stuck on this point, so I thought I’d take a look at the academic evidence on the subject; here is what I discovered: Cell datasheets only ever publish a 1kHz impedance measurement for internal resistance (IR), but unfortunately, this figure is not the same as a cell’s DC IR. A 1kHz impedance measurement is useful for predicting a cell’s performance when driving high frequency inverter-drives (as in an electric car), and it’s also an easy measurement to make for rapid pass / fail quality checks; and that is why it’s in common use.

To estimate PSC, you need a cell’s DC IR, and from my investigations, it seems that this is typically about 25% greater than its1kHz IR measurement. So, it’s possible to estimate PRC from a cell’s datasheet as follows: A cell with a 1kHz IR of 300 micro-ohms will have a DC IR of about 375 micro-ohms. In order to use a fuse, there needs to be a minimal circuit of a cable plus two joints (one to the cell, and one to the fuse). Taking the cable resistance as zero ohms, and a minimum achievable joint resistance of, say, 50 micro-ohms gives a total circuit resistance of 475 micro-ohms. The cell PSC can be then be estimated as 3.2 V / 475 micro-ohms, or 6,736 A.

Putting cells in series doesn’t change PSC, but putting two cells in parallel doubles it to 13,472A (and four in parallel to 26,944A etc.); that is why high AIC fuses such as Class-T or NH are required by ABYC and ISO. A battery’s PSC depends a lot on how many cells are connected in parallel, so a 10kA AIC MRBF fuse might be ok for 1P4S batteries (just about), but ANL or Mega fuses are not suitable. Interestingly, 1,000Ah Winston Thundersky cells have a datasheet IR of 300 micro-ohms, so in theory a 12V (12.8kWh !) battery built with these cells could be fused with an MBRF fuse. In other words, PSC is not necessarily directly related to Ah capacity.

In fact, one of the reasons for building LFP banks with a small number of large cells, rather than paralleling multiple batteries (see Rod Collins and others), is to limit PSC, making this arrangement arguably safer. Other benefits include: simplicity, less need for cell balancing, fewer busbars and joints, less cabling, fewer BMSs and fuses, greater reliability and a more compact construction etc. I gather that EVE 600Ah cells will soon be available, so a 12V battery built with just four of these cells (plus a BMS of course) will probably be all the house battery that most UK boaters will ever need.
Click to expand...
Being a 24v boat, we need 8 cells in series to make a single 24v battery. We chose 280Ah cells. With two batteries in parallel, we have resilience should we have a battery/bms problem. We can comfortably run on a single battery until we fix a battery. We carry a spare bms as that is the most likely fault scenario.
Each battery has a Victron Smart shunt, isolator and NH fuse. Isolating a battery allows us to carry on as normal on a single battery whilst a bms is changed, etc. Each batteries is in a separate closed battery box with mechanical ventilation so working on one battery doesn't put the other at risk
 
The downside of batteries in parallel is that because of the inevitable wiring and MBS resistance variations (even a few milli-ohms), the batteries won’t charge and discharge equally. One battery will always get more use than the others (which is not good), and could lead to premature failure and/or cascade failure. For example, consider two 200A BMSs driving a 300A inverter. Imbalances could cause one to trip, which would shut down the other battery immediately.

A well-loved LFP cell is probably the most reliable thing on a boat; they simply don’t fail. Blue Sea contactor BMSs don’t fail either. My contactor BMS system does have a redundant BMS, but that only cost me about £20 (as I designed and built it myself).

I understand that everybody will have different ideas about optimising their boat’s systems, and I respect other people’s choices for redundant systems. My mantra has always been ‘make it simple (and as high performance) as possible’, because the more things you have on-board, the more there is to go wrong…
 
migs said:
The downside of batteries in parallel is that because of the inevitable wiring and MBS resistance variations (even a few milli-ohms), the batteries won’t charge and discharge equally. One battery will always get more use than the others (which is not good), and could lead to premature failure and/or cascade failure. For example, consider two 200A BMSs driving a 300A inverter. Imbalances could cause one to trip, which would shut down the other battery immediately.

A well-loved LFP cell is probably the most reliable thing on a boat; they simply don’t fail. Blue Sea contactor BMSs don’t fail either. My contactor BMS system does have a redundant BMS, but that only cost me about £20 (as I designed and built it myself).

I understand that everybody will have different ideas about optimising their boat’s systems, and I respect other people’s choices for redundant systems. My mantra has always been ‘make it simple (and as high performance) as possible’, because the more things you have on-board, the more there is to go wrong…
Click to expand...
A Jk 200A BMS with 2A active balancer is circa £85.
Running twin batteries in parallel is really not a problem. With independent Victron smart shunt on each battery, I can see how close each battery is to the other. They really don't move that far apart. The fact that they are operating at the same voltage helps share the load more evenly. You can even run different sized batteries quite succefully. I did that for a season. 105Ah 24v battery in parallel with 280Ah battery. 3000w inverter running 2000w load on a regular basis. Induction hob, watermaker, immersion heater, etc we never had an outage. We never had an issue of one battery doing anything other than sharing its load proportionally to its rating. There is no way a small resistance difference in two similar sized batteries is going to see one battery taking such a high load or even varying by more than a few amps. My experience confirms it simply doesn't happen
 
geem said:
Being a 24v boat, we need 8 cells in series to make a single 24v battery. We chose 280Ah cells. With two batteries in parallel, we have resilience should we have a battery/bms problem. We can comfortably run on a single battery until we fix a battery. We carry a spare bms as that is the most likely fault scenario.
Each battery has a Victron Smart shunt, isolator and NH fuse. Isolating a battery allows us to carry on as normal on a single battery whilst a bms is changed, etc. Each batteries is in a separate closed battery box with mechanical ventilation so working on one battery doesn't put the other at risk
Click to expand...
Precisely the same architecture and gear of our system, with the same reasoning behind it, with the only exception that we are using T-Class rather than NH fuses (and I think NH are better actually).
 
geem said:
A Jk 200A BMS with 2A active balancer is circa £85.
Running twin batteries in parallel is really not a problem. With independent Victron smart shunt on each battery, I can see how close each battery is to the other. They really don't move that far apart. The fact that they are operating at the same voltage helps share the load more evenly. You can even run different sized batteries quite succefully. I did that for a season. 105Ah 24v battery in parallel with 280Ah battery. 3000w inverter running 2000w load on a regular basis. Induction hob, watermaker, immersion heater, etc we never had an outage. We never had an issue of one battery doing anything other than sharing its load proportionally to its rating. There is no way a small resistance difference in two similar sized batteries is going to see one battery taking such a high load or even varying by more than a few amps. My experience confirms it simply doesn't happen
Click to expand...
Indeed, but that doesn't mean you shouldn't take care in getting the cabling exactly right. Our two batteries are paralleled right inside the battery box with precisely measured identical cables.

The active balancer should take of any minor variations.
 
Dockhead said:
Indeed, but that doesn't mean you shouldn't take care in getting the cabling exactly right. Our two batteries are paralleled right inside the battery box with precisely measured identical cables.

The active balancer should take of any minor variations.
Click to expand...
My cables are the same length on my pair of 280Ah 24v batteries. They weren't when I had two different size batteries but it worked fine. I think the theoretical requirements and the actual requirements are a little different. The balancer will look after an individual battery. Whether one battery is slightly more charged than the other could be a function of how long you have as an absorption period. I found it didn't really matter. Lithium is so far advanced than lead that a reduction in cycle life is irrelevant. They will die of calender aging first
 
geem said:
My cables are the same length on my pair of 280Ah 24v batteries. They weren't when I had two different size batteries but it worked fine. I think the theoretical requirements and the actual requirements are a little different. The balancer will look after an individual battery. Whether one battery is slightly more charged than the other could be a function of how long you have as an absorption period. I found it didn't really matter. Lithium is so far advanced than lead that a reduction in cycle life is irrelevant. They will die of calender aging first
Click to expand...
I agree about cycle life. Mine should outlive me, I should think, or at least age out long before they cycle out, and plus or minus even a thousand cycles won't be noticeable. It's hard to get used the profound difference to lead.

Paralleling batteries with unlike cables won't just reduce cycle life somewhat but also capacity. Maybe also not highly relevant; YMMV.
 
Dockhead said:
Paralleling batteries with unlike cables won't just reduce cycle life somewhat but also capacity. Maybe also not highly relevant; YMMV.
Click to expand...
I would be interested to know why you think this?
If you charge at high current, maybe a large alternator direct to lithium or a diesel generator with large charging capacity, then the resistance in the battery cables becomes relevant. The higher the current, the higher the impact of unequal cable lengths will be. You may see a different voltage at each battery. Each battery doesn't know its in parallel with another battery with a slightly different voltage. The voltage difference only occurs whilst the charge is being applied. Take the charge source away and the voltages will equalise. So one battery may be slightly more charged than the other.
Since lithium doesn't need to be taken to 100% charged on a regular basis and you dont need to push cell voltage higher than 3.5v, what's the downside?
If you predominantly charge with solar, the current going into the batteries is likely to be far less than that seen with large alternators. The difference in battery voltage will be far smaller.
I think where it may be a problem is if you have 2 fairly small capacity batteries with unequal cables lengths, and you are charging them to their upper voltage limits to get the most charge into the batteries. It is possible that the battery with the higher charge state trips the bms on over voltage protection. This in its self won't damage anything but being a those upper limits may stress the battery.
When I ran two different size lithium batteries, with unequal cable lengths, the largest voltage differential I saw was 0.2v when charging with the diesel generator.
On my 24v system, I charge the batteries to 28v, so 3.5v per cell. In theory the 0.2v difference could cause one battery to trip but I only ever do generator charging if the batteries are low. I never need to use it to get the batteries to 100% charged. This is where solar comes in and that would be at lower current.
Just for fun, I might swap one of my battery cables for a longer one next week and see what happens with two equal size batteries. I suspect I could create a voltage difference between the batteries but does it matter?
 
I'm not sure equalising cable resistance matters much. I think it's because the batteries and the wires (should) have such low resistance they can accept current and so don't drop voltage significantly at the low currents that will occur when balancing happens That's my experience: I've got two lithium battery banks (port and starboard) and each fed with its own wire from the chargers. Port has a 100A charger and stbd has a 200A charger. Batteries are paralleled also with separate wires - it's an inerited installation converted from previously two separate lead battery banks and now joined up. A little messy I admit and with very different current flows and resistance values into and between the two banks, yet all the batteries are in perfect balance. 00.01v different if even that.
 
geem said:
I would be interested to know why you think this?
If you charge at high current, maybe a large alternator direct to lithium or a diesel generator with large charging capacity, then the resistance in the battery cables becomes relevant. The higher the current, the higher the impact of unequal cable lengths will be. You may see a different voltage at each battery. Each battery doesn't know its in parallel with another battery with a slightly different voltage. The voltage difference only occurs whilst the charge is being applied. Take the charge source away and the voltages will equalise. So one battery may be slightly more charged than the other.
Since lithium doesn't need to be taken to 100% charged on a regular basis and you dont need to push cell voltage higher than 3.5v, what's the downside?
If you predominantly charge with solar, the current going into the batteries is likely to be far less than that seen with large alternators. The difference in battery voltage will be far smaller.
I think where it may be a problem is if you have 2 fairly small capacity batteries with unequal cables lengths, and you are charging them to their upper voltage limits to get the most charge into the batteries. It is possible that the battery with the higher charge state trips the bms on over voltage protection. This in its self won't damage anything but being a those upper limits may stress the battery.
When I ran two different size lithium batteries, with unequal cable lengths, the largest voltage differential I saw was 0.2v when charging with the diesel generator.
On my 24v system, I charge the batteries to 28v, so 3.5v per cell. In theory the 0.2v difference could cause one battery to trip but I only ever do generator charging if the batteries are low. I never need to use it to get the batteries to 100% charged. This is where solar comes in and that would be at lower current.
Just for fun, I might swap one of my battery cables for a longer one next week and see what happens with two equal size batteries. I suspect I could create a voltage difference between the batteries but does it matter?
Click to expand...
This would indicate that it doesn't matter where on the parallel bank you take the positive or negative leads from as any difference will be balanced eventually.

That's my OP question answered. Thanks 🙂
 
mattonthesea said:
This would indicate that it doesn't matter where on the parallel bank you take the positive or negative leads from as any difference will be balanced eventually.

That's my OP question answered. Thanks 🙂
Click to expand...
I think it's different with lead to lithium. Lead needs to reach full charge to stop sulphation. Lithium has no such issue. My lithium batteries are both at 26.2v this morning. If they had slightly different levels of charge, it really doesn't matter. My shunts suggest there is 4% difference in their state of charge. What i find is that depending on their charge state one can be taking more load than the other. Unless you have every cell picked to have identical charge profiles, there will be differences in the way the cells accept charge and release it. It doesn't seem to matter
 
Following this discussion I’m glad I have one 460 aH Lifepo4 battery!

However I agree with everything said about the benefits. With generator and engine on and solar charging etc I’ve seen over 120 Amps going into our battery. (We have 50 amp DC-DC off the engine and an Inverter/charger that puts out 80 amps when you apply 230 AC to it.) It doesn’t take long to be up to near 100%! Mostly I don’t bother as our 430 watts of Solar keeps up with most things at all times. It’s also nice not to always have to run the generator to run the watermaker.

One trick from a friend is to use a thermal imaging camera to go round and look at all the joints you’ve made. Any joint showing slightly high resistance is obvious on the thermal imaging camera.
 
PaulRainbow said:
Those figures are for AC, DC will be much, much less.
Click to expand...
There is a derating. About half on the volts, but that is not an issue. With current rating is much less clear. I was hoping there might be a table or calculation, but I found little info. According to some technical sheets there is between zero (RS Pro) and an 80% derating. There is almost certainly headroom. The specs vary wildly and it would be useful to have more clarity.
 
The difference in rating between AC and DC for a component can be very large.
I can't help with the fuses, but as an example I bought a switch rated for 20A AC and discovered that it was only rated for 1A DC.
 
You must log in or register to reply here.

Other threads that may be of interest

Baggywrinkle
Test run of the Lithium system that has been sitting in my cellar for most of this year.
Replies
5
Views
311
claymore
claymore

Share this page

Top