An Affordable Lithium Boat Battery

[geem]

geem; understood, we should coin a new word for use with lithium batteries instead of float - let's call it 'hold' voltage.

In fact, it's actually impossible to define a best hold voltage for Lithium as the charge going in and out of the battery varies with the load applied. If you set the hold voltage high, that's fine for powering big loads and maintaining soc, but when the loads are light e.g. you are away for a few days, the battery gets overcharged. If you set the hold voltage low, the opposite happens.

So, with a fixed hold voltage, you inevitably end up constantly charging or discharging the battery, for example as the AC cuts in and out, and after a while you end up with an indeterminate soc. For us a better approach was to ensure that charge flows in and out of the battery in a controlled fashion, and this is done by disconnecting the battery from the chargers when it has charged i.e. no bulk, absorption or float stages at all, just simple constant current charging. (A side benefit of this approach was the possibility of retaining our old chargers and saving around £800.)

Maybe we should have taken out a patent on this novel approach, but surely we can't be the first people to come up with it?

We are full-time liveaboard. We don't leave the boat for a few days. We use a Victron smart shunt so we can measure where we are for SOC. We are counting amps in and out of the battery so no problem working out SOC.
Having float set at 3.35v/cell simply ensures the lithium comes off charge once its hit the threshold charge voltage level ( with a 30min absorption period to assist balancing). From what I am aware, this is normal practice and the same as what you do., accept you don't do any cell balancing. With your small loads, there is probably no reason to balance. We are regularly running loads of 1800w. Cells will go out of balance so a good active balance process is sensible.

We do have an old style charger that we can run via the diesel generator but we never run it to do a full charge cycle. It simply bangs in 30A whilst we are running the watermaker if, we have a cloudy day. This has happened about once per month so far.
If I was leaving the boat for a period, everything would be switched off and the batteries would be left at approximately 50% charged, depending on how long we plan to leave the boat.

 

[migs]

'Without the buffer of the battery, the voltage fluctuations can damage sensitive electronics.'

That's not how our system works. Loads remain connected to the battery, but the chargers are able to be disconnected from the charge buss.

Do you worry if a plugged in mains charger is not connected to a battery? It's the same with mppts. There's something seriously wrong with their design if that is the case - and I've designed a few myself. Incidentally, we don't use Victron as it would not perform well with our panel. Is it Victron mppt controllers that have this defect?

 

[noelex]

Do you worry if a plugged in mains charger is not connected to a battery? It's the same with mppts. There's something seriously wrong with their design if that is the case - and I've designed a few myself. Incidentally, we don't use Victron as it would not perform well with our panel. Is it Victron mppt controllers that have this defect?

Some mains battery chargers have the facility to be a power supply. These have more buffering on the output and can be left connected to power systems (providing the output is sufficient) without a connected battery.

Unfortunately, many mains chargers (and no MPPT solar controllers that I am aware of) can do this. It will be listed in the specifications.

Disconnecting the battery will not usually damage a mains charger, but unless it has a "power supply" option any sensitive connected electronics may be damaged.

 

[migs]

'We are regularly running loads of 1800w. Cells will go out of balance so a good active balance process is sensible'

We have a 4.8kW bowthruster load, but still normally have <2mV difference between cells in over two years without ever doing any BMS balancing. Perhaps the high but short load duration is the reason?

 

[migs]

Our mppt controller has a buck/boost synchronous architecture that improves performance with our panels compared to commercial controllers. When designing it, I didn't even consider battery removal issues as this sort of safety protection should surely be a given. I'm quite shocked to hear that all commercial mppt controllers suffer from this glaring defect.

 

[migs]

I wonder if we are at cross purposes with the mppt thing:

I thought people were concerned about damage to the mppt controller if it's battery was disconnected, but perhaps you believe that we need it to act as a power supply. That's not so with our system as in #62 I said:

'That's not how our system works. Loads remain connected to the battery, but the chargers are able to be disconnected from the charge buss'

In other words, when the battery has charged, only the battery is connected to house loads and the chargers are effectively open circuit.

 

[geem]

We prefer to let our battery control its own charge regime rather than have multiple chargers (with multiple settings) fighting it out.

No mppt controller should ever be embarrassed by disconnecting its battery. What sort of design would that be? If they were, Victron would surely be knee-deep in returns.

That's not how it works. The victron mppt will be collecting accurate battery voltage from the smart shunt.

'We are regularly running loads of 1800w. Cells will go out of balance so a good active balance process is sensible'

We have a 4.8kW bowthruster load, but still normally have <2mV difference between cells in over two years without ever doing any BMS balancing. Perhaps the high but short load duration is the reason?

You won't know how much the cell difference is until you raise the charge voltage. If you are on the flat part of the curve, you won't see the cell imbalance. That's why you don't turn balancers on until you reach 4.45v/cell or higher

 

[geem]

I wonder if we are at cross purposes with the mppt thing:

I thought people were concerned about damage to the mppt controller if it's battery was disconnected, but perhaps you believe that we need it to act as a power supply. That's not so with our system as in #62 I said:

'That's not how our system works. Loads remain connected to the battery, but the chargers are able to be disconnected from the charge buss'

In other words, when the battery has charged, only the battery is connected to house loads and the chargers are effectively open circuit.

Victron mppts only need to be connected to a battery initially to recognise the battery voltage. After that, there is not a problem with having pv connected but no battery.
If you connect the smart shunt after the BMS and the BMS creates a shutdown, the smart shunt sees the zero battery voltage and shuts down the mppt

 

[noelex]

I thought people were concerned about damage to the mppt controller if it's battery was disconnected

Yes, that is a concern.

Generally, it is better to disconnect the solar input before disconnecting the battery supply to avoid damage to the controller, although the risk of damage is not high.

'That's not how our system works. Loads remain connected to the battery, but the chargers are able to be disconnected from the charge buss'

The second and greater concern is damaging any connected (to the solar controller) sensitive electronics. This depends on how your system is wired. On some boats it is not possible to power the systems via the solar controller without a connected battery in which case this particular problem can be ignored.

 

[migs]

'You won't know how much the cell difference is until you raise the charge voltage. If you are on the flat part of the curve, you won't see the cell imbalance. That's why you don't turn balancers on until you reach 4.45v/cell or higher'

Surely its better to have a system that runs without cell differences. There is no need to push cells into the danger zone.

'The victron mppt will be collecting accurate battery voltage from the smart shunt'

When connected, why can't the Victron mppt collect an accurate battery voltage from the battery. When disconnected it doesn't care.

 

[geem]

'You won't know how much the cell difference is until you raise the charge voltage. If you are on the flat part of the curve, you won't see the cell imbalance. That's why you don't turn balancers on until you reach 4.45v/cell or higher'

Surely its better to have a system that runs without cell differences. There is no need to push cells into the danger zone.

'The victron mppt will be collecting accurate battery voltage from the smart shunt'

When connected, why can't the Victron mppt collect an accurate battery voltage from the battery. When disconnected it doesn't care.

The shunt is located within 300mm of the battery. Not so for our MPPTSs. The shunt is more accurate, that why Victron use it.

 

[migs]

'The second and greater concern is damaging any connected (to the solar controller) sensitive electronics'

We trust that you are familiar with the concept of separate charge and load busses (see the article if not). In our system its physically impossible for the mppt controller to connect to boat loads without the battery in circuit.

'The shunt is located within 300mm of the battery. Not so for our MPPTSs. The shunt is more accurate, that why Victron use it.'

Understood; our Blue Sea systems monitor does the same.

 

[geem]

'The second and greater concern is damaging any connected (to the solar controller) sensitive electronics'

We trust that you are familiar with the concept of separate charge and load busses (see the article if not). In our system its physically impossible for the mppt controller to connect to boat loads without the battery in circuit.

'The shunt is located within 300mm of the battery. Not so for our MPPTSs. The shunt is more accurate, that why Victron use it.'

Understood; our Blue Sea systems monitor does the same.

Yes, Victron, EV Power and Heltec to name a few do contactor controlled BMS similar to yours.
There are pros and cons with FET bases system and contactor based system. If both are well designed then in my opinion, there is little to choose between them.
They have strengths and weaknesses that good design should take into account

 

[noelex]

We trust that you are familiar with the concept of separate charge and load busses (see the article if not). In our system its physically impossible for the mppt controller to connect to boat loads without the battery in circuit.

'

I understand that in your setup it is not possible to power systems (such as marine instruments) without a connection to the battery, but this does not apply to many boats, therefore it is a risk that needs to be considered, especially with lithium batteries where the BMS may automatically disconnect the battery without user input.

Most people are aware of the possible damage to the alternator in the event of a battery disconnect and take steps to avoid this potential damage, but many do not consider damage to other electronics if the solar system (or other charge sources) can remain powering the house system.

 

[gregcope]

A couple of years ago we looked into upgrading our failing AGM house batteries to lithium, but unfortunately the price of a commercial system was somewhat beyond our budget. Nevertheless, we wondered whether a diy approach would bring the cost down to an affordable level and started investigating. Published designs didn’t seem to meet our needs, but by taking a fresh approach and challenging some received wisdom we came up with our own design; an affordable solution that works well. We’ve been enjoying the benefits of Lithium for more than two years now, and our system has proved both reliable and effective. Here is the story; we hope it is of interest. bit.ly/43D3X1h

Nice write up @migs. Got any more details on your BMS and relay/contactor used?

 

[noelex]

When connected, why can't the Victron mppt collect an accurate battery voltage from the battery. When disconnected it doesn't care.

Without any accessories such as a Victron battery monitor or a Victron Smart battery Sense device, a Victron solar controller will rely on controlling its output voltage. This is also the case for most non Victron systems.

If there is no battery in the system the controller will still attempt to maintain a steady output voltage (either the bulk/absorption or float set point). This will be powering the house system even if the battery is disconnected in many installations.

If a high load switches off then the output voltage will spike above the voltage set point. This happens even if there is a battery in the system. The solar controller cannot adjust fast enough to keep the voltage absolutely constant. This spike is minor and is not an issue, but without the buffering effects of the battery the voltage spike is many times higher. Without a battery It can, and sometimes does, reach the stage where sensitive electronic systems (such as boat instruments) are destroyed because their maximum input voltage is exceeded. This can be very expensive.

This is not a critisism of your system. I understand this is not possible with your configuration, but it is a common issue that should be considered. It is possible to protect against this fault as long as people are aware of the problem.

 

[Zing]

Lead and lithium batteries are fundementaly very different. One advantage of lead is that it can deliver very high current for a short period of time without any damage. As well as differences in the battery chemistry with lead there is no limitations from the BMS (as these are not fitted).

Even deep cycle lead acid batteries can typically deliver well over 5C for thirty seconds without damage, although they will suffer significant voltage sag.

This means that for very high draw devices you need to be careful simply switching lead acid for lithium without doing some homework. This is the primary reason why lithium batteries are not commonly used for engine starting in marine situations. Bow thruster operation driven by lithium is feasible, but some care is needed to make to make sure the current delivery is adequate, especially when the limited capabilities of some BMSs are considered.

I have a high peak load system, so picked a BMS that has no current limit. It only cuts out for high charge volts and low discharge volts, not on amps. I can feed hundreds of 27v amps into my engine starter or bow thruster, furlers etc all at once if I want and it works perfectly. For my boat it’s the correct solution.

 

[noelex]

I have a high peak load system, so picked a BMS that has no current limit. It only cuts out for high charge volts and low discharge volts, not on amps. I can feed hundreds of 27v amps into my engine starter or bow thruster, furlers etc all at once if I want and it works perfectly. For my boat it’s the correct solution.

All BMS systems will have a maximium current limit.

The higher power BMS units will send an impulse to a heavy duty relay/contactor that does the actual switching , but this still has a finite current limit. If the load on the house battery is higher than than the heavy duty relay/contactor can conduct (or interupt) the system will be damaged.

 

[Zing]

All BMS systems will have a current limit.

The higher power BMS units will send an impulse to a heavy duty relay/contactor that does the actual switching , but this still has a finite current limit.

No. Not all have a current limit.

 

[noelex]

No. Not all have a current limit.

If you have a high powered BMS it will have a heavy duty relay/contactor as part of the BMS system. If you look up the specifications for this device, or these devices (sometimes more than one can be used) it will have a maximum current value. The BMS will not limit the maximum current, but ithe contactor will be damaged if its current limits are exceeded. Therefore the BMS sytem has a limited maximum current.

Common used marine contactors are pictured below, although there are many choices: