32s LiFePo4 with 250A MOSFET BMS

[maB#02]

I am pretty sure I know the question to this answer, but I want to make sure nonetheless.

I want to change the batteries of a Piaggio porter electric from lead gel to 32 3.2V LiFePo4 cells. It has a 96V motor, rated at 11kW nominal with up to 22kW for up to 60 seconds.
In this video of a similar project, at 2:30 you can see that the drawn current goes up to 400A, although only up to a very short burst (around 3 seconds).

At 22kW and 96V the motor should draw only 230A max, but I guess the inverter draws more and some is lost during conversion (400A aat 96V = 38.4kW with 22kW motor → ~60% efficiency at that power draw?)

At first I was eyeing a Daly 32s 250A BMS, because I thought 230A should be enough for 22kW. But looking at the video, obviously this is not the case.
Still, most of the time I would be within the 250A specification.
The sales representative (... I know ...) of the Daly BMS said short bursts of 400A should be okay for the 250A BMS.
From my technical understanding, the main problem is the heat dissipation in the MOSFETs. Even though the BMS has a fan, I am afraid this will fry the MOSFETs.
Am I correct in this assumption?

 

[GXMnow]

To know for sure, I would need to find the full specs for the MOSFETs they are using. But I do know many MOSFETs can handle some crazy over current in short bursts. It does come down to heat in the channel. It takes time to get hot, and it is a balance act with the cooling system.

That being said... I would not run the motor current through the MOSFETs. Have the output of the BMS power a contactor that connects the motor inverter directly to the battery bank. The BMS will not see any of the motor draw or regen current, so the SoC calculation will be way off. But the BMS will only need to deal with your DC to DC converter and other minor vehicle loads. You could probably get away with a 100 amp BMS then. If the BMS goes into a fault condition, the contactor will still drop out and protect the battery from cell or pack over discharge or charge. The only thing it won't protect for is over current. The motor controller should have current limiting. That should be set to the maximum C-Rate of the cells, or a little less. I would then also add a 500 amp or higher "Smart Shunt" to measure the full current in and out of the battery to give you a better SoC estimate.

 

[maB#02]

Thank you! That sounds very interesting, could you give me a very raw schematic of how this would look?

The only thing it won't protect for is over current. The motor controller should have current limiting. That should be set to the maximum C-Rate of the cells, or a little less.

I am using EVE LF304, they have a maximum continuous discharge rating of 1.0C (=304A or a little less when the batteries degrade later on), but I assumed that going to 1.3C for 3 seconds shouldn't be too much of a problem?

 

[GXMnow]

I don't have a decent drawing tol on this PC, so I will try to describe it.

The battery cells all connect to the BMS as expected. The MOSFETs typically will break the negative end of the pack. The output side of the BMS will drive your DC to DC converter that keeps your normal car loads running and the 12 volt battery charged. That should be turned on and off by the key so it only draws when the car is on. If the BMS goes into any shut down, the DC to DC will shut off, but all of the car loads like the headlights will still be able to run off the 12 volt system battery.

From the BMS output, use a small DC to DC buck converter down to 12 volts to the ignition switch. This will be hot when the BMS says the HV battery is good. Then have this switched 12 volt source power a 500 amp contactor to connect the battery pack negative before the BMS to the motor inverter/controller. If the BMS goes into shut down for any reason, that buck converter will also shut off. You lose the 12 volt ignition feed, and the contactor disconnects the motor from the battery.

If you only have a single input ignition switch, you might want to add a second 12 volt powered relay. The ignition switch powered by the 12 volt battery would then turn on this new relay. That relay connects the small DC to DC buck converter to the 96 volt battery bank after the BMS. That converter then powers the 500 amp contactor. This way, the other items powered from the ignition switch will still have 12 volt battery power. But if the BMS shuts off, the 500 amp contactor still shuts off.

Let me know if I am not clear on all of it. I can try drawing something up on my work PC, it has Adobe Acrobat with drawing tools.

Will did a video about a small home solar build where he used a very small BMS to control a contactor. It's the same basic idea.

 

[maB#02]

Okay, here's a rough sketch of my understanding. Am I correct?
I have omitted the 96V to 12V buck converter, because I found this power contactor that can take 96V on the coil circuit:

EV200ADANA DC-Hochstrom-Relais, Spule 32V-96V DC, Hauptschütz, DC-Contactor 500A, 900VDC, 30mA Haltestrom @48V

Leichtes, kleines, vergossenes DC Lastrelais, Hochstromrelais, Contactor, Schüzt für Anwendungen der Elektromobilität sowie als Sicherheits-Trenn-Relais zwischen Batterie und Wechselrichter, Motorcontroller... EV200ADANA 32V-96V ...

www.nothnagel-marine.de

To be honest, I haven't understood the part with the second relay.

 

[yabert]

32 3.2V LiFePo4 cells

Can I suggest you yo look at module coming from EV.
The energy density is way better, prices are right and quality is at top.

The EV200 contactor is right for your project, but you need to control it with 12V, not 96V.

 

[GXMnow]

To be honest, I haven't understood the part with the second relay.

The EV200 contactor is right for your project, but you need to control it with 12V, not 96V.

Responding to both previous posts...

In your drawing, where you have the switch labeled "Ignition", make that a 12 volt relay. Then the coil will be controlled by the vehicle normal 12 volt ignition switch. That way, your 12 volt systems all work as original for a 12 volt vehicle, and the key is not trying to switch 96 volts. And you still have all the protection as the feed for the high voltage contactor will drop out and shut it down if the BMS sees any fault. That contactor looks good, but I think the coil is actually just 48 volt. The contacts are rated to break 96 volts.

Also keep in mind, LFP cells are 3.2 volts nominal, 2.5 volts at zero SoC and full absorb voltage is 3.65 volts.
So the 32S LFP battery will have a nominal voltage of 102.4 volts, and fully charged to 100% could hit 116.8 volts. But it will settle back to 110 volts after the charge current is removed. If you do run it to dead, the voltage can fall as low as 80 volts. Does your system have a "Pre Charge" setup? Usually an EV will have a second contactor that connects through a resistor to slow the inrush current to charge the capacitors in the motor controller. Then after a fraction of a second, it closes the main contactor to carry the power. This reduces stress on the caps as well as greatly extending the life of the contactors. Even my hybrid does this, I can here the click... CLICK when I turn the car on.

 

[maB#02]

The contactor I linked is the EV200ADANA, that should have a coil operating voltage of 32-95DC and be able to handle a power side voltage of 12-900V, right? See image below.

Although I am not sure if it can handle the 110V of a fully charged 32s then.

Does your system have a "Pre Charge" setup? Usually an EV will have a second contactor that connects through a resistor to slow the inrush current to charge the capacitors in the motor controller. Then after a fraction of a second, it closes the main contactor to carry the power. This reduces stress on the caps as well as greatly extending the life of the contactors. Even my hybrid does this, I can here the click... CLICK when I turn the car on.

I don't know that currently for the Piaggio Porter Electric but I'll try to find out.

Can I suggest you yo look at module coming from EV.
The energy density is way better, prices are right and quality is at top.

For my use case the 31kWh of the 32s EVE LF304 are enough. The Porter Electric is not a strong or fast EV, I'll use it mostly for local transport.

 

[yabert]

The contactor I linked is the EV200ADANA, that should have a coil operating voltage of 32-95DC and be able to handle a power side voltage of 12-900V, right? See image below.

Right.

Still, you need the EV200 with the 9-36V coil to control it with the basic 12V system of the vehicle.
Don't know about shipping cost to europe, but really affordable here: https://batteryhookup.com/products/te-connectivity-ev200aaana-500a-0-900vdc

 

[maB#02]

Right.

Still, you need the EV200 with the 9-36V coil to control it with the basic 12V system of the vehicle.
Don't know about shipping cost to europe, but really affordable here: https://batteryhookup.com/products/te-connectivity-ev200aaana-500a-0-900vdc

In my (kindergarden-) schematic, there are two "switches", the one labelled contactor, and the ignition switch (which should really also be a relay/contactor like GXMnow said).

I agree that the ignition labelled relay/contactor should have 12V coil, but the "contactor" labelled contactor's coil is connected directly to the 32s batteries, so 96V. Is there an advantage to convert 96V in that circuit to 12V and use a 12V coil contactor there? I guess one advantage would be that the "power" side of the ignition relay could also be chosen for only 12V, but the contactor in the link you posted is very cheap anyways.