Low Temperature Disconnect Behavior

[Philip Schuchardt]

Yesterday morning I woke up and noticed that the battery temperature from the ExtTemp was at 4.9C. It was interesting because a charge was still being accepted from the PV. I thought that SBMS0 would disconnect charging below 5.0C. The temperature eventually got down to 4.7C before warming up above 5.0C. I started reading the documentation again, and realized that the min temp needs to be set at 3000, instead of 4092 (the default). So I set the temp to 3000 (5C), saved the parameters, but didn't reset the SBMS0 (by removing balance cables).

Last night, the outside temps were around -10C. The internal van temperature was about 21C. Around 5:58AM, the all power cutoff. Below is a screenshot 1. I assumed that the temp dropped above 3000 (below 5C). In my case cutting off discharge is really bad, because it turns off the van's heater. This behavior will make the batteries even colder. I turned the van on, and blasted the heat for about 30 minutes warming the batteries up. The batteries then were warmed up below 3000 (above 5C). Unfortunately, everything was still off. In an effort, to get the heater going again, I set the low temp to 3060 (~2.5C), still nothing. Then I reset the SBMS0 by unplugging the balance cables. After the reset, everything turned back on (screenshot below 2). 

1. Before reset

2. After reset, and min temp at 3060 (~2.5C).

Questions:

1. If the Ext temp drops below the min temp, does this turn both charging and discharging off? 

2. If the batteries go below the Min temp (for example <5C) and everything gets disconnects, when does it become reconnected? Does it have to rise from 5C to 7C or something (didn't see any documentation on this)? 

3. In the SBMS0 manual, it states that batteries need to be above 5C for charging, most LiFePo4 battery specs say 0C. What's the reasoning behind the difference? 

4. Clearly my batteries need a heating pad with thermostat. They're in a fairly insulated box since they stay cold. I have 300A at 12V 4S2P. What size heating pad in watts do I need? Any recommendations on thermostats? 

Thanks,

Phi|ip

[Dacian Todea]

Philip,

1. Yes that is related to the ISL94203 and can not be modified so both charge and discharge is stopped if either the low or high battery limits are exceeded.

2. There is a certain hysteresis not sure exactly what it is but yes once the temperature increases it will be reconnected.

3. Yes above 0C is in most spec's I just like to be conservative and recommend +5C or even +10C

4. If the battery is in a proper insulated box a 15 to 20W heater with thermostat set at +10C should be more than sufficient even if there is a -20C average outside.

Any thermostat should work maybe one of those low cost digital thermostat controllers  https://electrodacus.com/solarforum.html?place=msg%2Felectrodacus%2FKIXOkeAMImc%2F27qRzPs4BgAJ

[Philip Schuchardt]

Dacian,

Thanks for the response! I'll probably be ordering a heating pad and thermostat soon. I agree that 5C is definitely a good rule a thumb since you can't guarantee that the whole pack is >0C, but just the temperature sensor. I'm sure in my case, the bottom of the pack is cooler than the top of the pack, since most of the cold air in my van comes from windows or the floor. The temperature sensor sits between 2 of the cell at the top of the pack. 

Phi|ip 

[Peter]

I have used 4x30W pads (15€ together) and put them between every second cell, but connected them in two parallel strings with two in series, so its still a total of 30W heating power but nicely distributed. This seems to work good, with the cells aluminum shell working as heat piping. I have one sensor on both sides and have a maximum of 1-2C difference between them. 

With my less than favorable insulation it takes roughly 10W average to have it 10C above ambient. Its 15mm of wood all around plus proper insulation only on some of the surfaces, because of space limitation.

I used the white temperature controller Dacian has tested, because it was quickly available but I also ordered another one, because of power draw and I am not sure how safe it is for up to the 27V of the LFP considering Dacians comments on the internals. I does seem to work well sofar though.

[Peter]

So I just measured out the hysteresis. It seems to be exactly 100 on the raw value, around 3K. 

@Dacian: I suppose there is no way of changing this value?

[Dacian Todea]

Peter,

Yes the hysteresis is done in hardware and needed else you will have ON/OFF cycles where you are around the trigger temperature and that as you measure is around  3C or 3K same thing.

The important thing will be to find a way to maintain the battery temperature above that limit whatever that is +5C or +2C is still a limit that you should never get to but it seems the battery is not properly insulated to prevent that from happening.

Of course you can bypass that by either heating just the temperature sensor or adding a say 20kOhm or so in parallel with the temperature sensor to trick the SBMS0 in thinking battery is much warmer than it actually is and so allow it to work even if you may damage to battery if is to cold. The damage is gradual if you charge when battery is below freezing still over time that will add up.

So the hysteresis is needed and eve if you were able to change that it will be of no help to you.

[Rob Duncan]

My system also shut down, both charge and discharge. I am using SBMS0 and all Victron equipment with the CCGX is connected to their VRM web portal via WiFi. This allowed easy temperature tracking although the Victron Battery Monitor temperature sensor, seen via VRM portal, is mounted to the battery post, while the SBMS0 is using the internal Tesla battery thermistors. They read about 1 degree C different.

SBMS0 ET reading has risen from 3C to 6C as the heated water begins flowing through the batteries which still show ~24V. I previously installed a "Enable Discharge Override" bypass switch on my 24 to 12 V converter. This allow s me to get heater and fluid pump operational. However the DFET is still off. I assume I can temporarily change the 5C (3000b) low temp shut-down value to allow charging again, but I also want to know at what temperature will the SBMS0 turn on the Charge and Discharge FETs?

My design pretty much permanently shuts down the system once the low temperature limit is reached. It would be better if I used any remaining battery power to continue attempt to heat the battery. It might be only a few hours away from sunrise when is shuts down due to low temperature limit. The fluid heater and pump only draw 50W. 

If I change the heating system to be fully operational with 24V direct from batteries, and not be controlled via the SBMS0 DFET, is battery discharge ok below freezing?       

[Dacian Todea]

Rob,

You will just use to much energy if battery is not thermally insulated. With an insulated battery you should be able to maintain battery temperature above +10C with no more than a 10 to 15W electric heater.

There is some hysteresis on the temperature (not sure exactly what that is as I did not used the temperature limit) but likely 1 or 2C

You can not just temporarily change that 3000b limit as to take effect you need to power cycle the SBMS0.  Any changes to the Parameter settings (or advanced parameter settings requires this as a safety measure).

So I will say set maybe the limit to +1C and then have your own thermostat set at 5C that disables the charging only if that is what you want.

But for anyone reading this the only proper solution when battery is not inside a thermally controlled environment and in a cold climate is to have the battery in a well insulated box and then add a small 10 to 15W heater (can be a power resistor + a small fan) controlled by a small electronic thermostat to maintain +10C in the battery box. As a backup you can add a bimetal thermostat to disconnect in case electronic thermostat fails.

 

[Rob Duncan]

Thank you yet again for your quick reply!

My exterior battery compartment already has 4 inches of blue construction foam board on all six sides. The exterior fluid lines are also insulated. I have tested system twice before while parked outside at -12C. During these tests it has never dropped below 10C, although it could not hold the usual fluid thermostat controlled programmed 20C. So this thermal shut down was very unusual instance, especially while parked within an insulated garage. The heating system generally cycles less than 50% duty cycle until about -5C. I am still investigating to determine the root cause, and have added more instrumentation.

You did not answer my two questions:

1. At what temperature do the FETs turn back on, assuming lower limit set at 5C?

2. What would be lowest temperature possible to draw about 50W from battery without significant damage?  

[Dacian Todea]

4" is a lot of thermal insulation so not sure why you are losing over 50W at less than 20C delta in temperature. My guess is that the fluid is where most of the losses are.  You should just remove that liquid heating and have just a 10 to 15W electric heater inside the battery box.

1. I think is +6C (as mentioned not sure exactly what the hysteresis on temperature is as I never used that other than for testing the software).

2. Drawing any amount of power is not a problem at any temperature just charging is the problem.

[Dacian Todea]

Just seen the battery photo and is hard to understand but it may look like you have a lot of thermal bridging so that may be the main reason for the huge heat loss.

[Rob Duncan]

I am not sure what thermal bridges you see. You can barely see the edge of blue foam on the far side of the farthest battery. About 1" of the blue foam is visible on the nearest edge. Four each 1/2" aluminum square tubing with heat shrink cushion is used for battery hold-downs that capture the battery mounting ridge. Red heat shrink for the nearer positive terminal is barely visible. There is no insulation between the batteries. Insulation on the shorter side is not seen, and the upper cover insulation has been removed so you can see the battery tops.

I tried to upload a couple more images, one showing the internal fluid line connections within the insulated box, and another of the two bare batteries as they were first inserted with no insulation electrical nor hydraulic connections. 

I kept getting "server rejected" on these 300 and 400 kb JPG files???     

[Rob Duncan]

I found that at 8C the SBMS0 allowed charging and discharging. I am using 5C limit.

[Peter]

I am repeating myself here: the hysteris is exactly 100 on the raw value. This is around 3,5K.

[Rob Duncan]

Sorry Peter, but I did not understand "raw value". Do you mean that if the lower limit is adjusted upward by 100, i.e. use 3000b + 100b = 3100, then the SBMS will then allow charging? Or are you saying to use 3500b? I simply shined a 200W floodlight on the battery box and noted when charging became enabled.

[Peter]

Almost: if you had 3000b set as low temp cut off, it will reconnect at 2900b which is the 8C you saw. 100b equals roughly 3,5 Kelvin which we use for temperature differences.
You can see the current raw value in diagnostic page.

I think that's something worth noting in the different manuals out there...

[Dacian Todea]

Rob,

It seems Peter already mentioned the hysteresis is 100 in raw value and seems to correspond to around 3 degree Kelvin or Celsius is the same thing.

What you say it is aluminium square tubing with heat shrink on it looked to me as steel tubing welded to the metal case.  Not sure how those aluminium tubes are mechanically connected to the case ?

Not sure what are the limitation for the jpg files on google groups.

[brianjonesphoto]

I have a question with this same issue. Right now my raw data reads 2985 for ext temp. I set the min ext temp to 2850b rebooted the SMBS and the discharge is still off. I took it down to 2700b and rebooted as well and the discharge is still off. If I understand this correctly the if the min is set 100 units under the current raw data the discharge should turn back on correct?

Thanks.

[Dacian Todea]

You are going the wrong way with the setting. That is a setting for the minimum temperature so while your reading is 2985 so around +5C you set the limit to 2700b that is close to +15C so charging will not start until you battery is way above +15C