MPPT Battery Charging - AC component during high PV loads

[CMDRZOD]

Hi there. Just wondering if anyone else has noticed this.
I'm running off grid, and using extra energy to run "loads of opportunity" when PV production is good.

My system:
4x Conext MPPT 100 600 (24kw charging, overpanneled at 32kw)
3x 6848 Pro Inverters (20.4kw sustained output)
6x AES Discover Rackmount (30.5kwh)

My system prioritzies battery charging, however I curtail charging to 0.2C (~120 Amps). (I do this to extend battery life) As battery temperature rises, I lower this charge rate until we hit 30C, at 10A.
As more charging is curtailed, loads of opportunity are started while the system attempts to maintain charging output at the set level, depending on battery temperature.

Thought excercise:
What I've noticed is that BMS reports 10A to batteries, my meter agrees when measuring DC amps. However at this point the inverters and solar output are nearly matched, 20kw PV in, 20kw output on inverters minus the 10-120 amps or so to batteries.

I switch my fluke true RMS meter to AC amps, and it will read 270-300 amps AC at exactly 120hz.

I also see battery temperature rise once these load of opportunity ramp up.

Any ideas on what I'm seeing? I believe its a combination of MPPT output and inverter draw causing this AC current to be apparent to my meter. In any case, I'm thinking I should have a capacitor to keep the batteries from having to deal with this AC high frequency current.

Would love to hear thought on this.

 

[wpns]

First question would be if the ripple current is even a problem. Second question would be what size capacitor you would need. https://www.powerelectronictips.com/choosing-capacitor-ripple-current-capability-matters-farads/ Has some thoughts but I’d imagine it’d be huge. In your case the battery is acting to smooth out the ripple at the expense of some internal heating across the ESR of the battery.
How much battery heating are you experiencing?

 

[CMDRZOD]

Hi there. I'm looking at an eaton 130 farad capacitor, rated for 62V part no XLM-62R1137A-R. The heating doesn't appear to be much, the highest reported temperature I've seen from the BMS is 34C which I don't believe is terrible, but not ideal. My goal is to keep battery temperature below 30C.

I believe this ripple, while not catastrophic, is detrimental to battery temperature and life.

 

[Vigo]

Im sorry, but just to be clear you are measuring this on the battery circuit, correct?

I cant imagine a capacitor would have better longevity than the battery bank if it were flowing 300a in and out 120 times a second. I guess id want to know whether that was true before i bothered trying to implement one. What sucks is its hard to assess the possible impact on battery life from this AC current.

 

[CMDRZOD]

Yes this is measured on the battery leads.

As far as cycling, I'd prefer the capacitor takes this current as opposed to the batteries.

Capacitor 1 million cycles
Battery: 10000 cycles

Yes I agree, I've never seen this talked about. I did not plan for it in my initial installation.

I used this simple calculator to calculate capacitor size required, for a given output.

 

[CMDRZOD]

Simple Energy Storage Capacitor Formulae

Went to:
I want to draw x amps for t seconds
Punched in:
270A
.0083sec (120hz)
54v
48v

According to that the capacitor size required is miniscule.

So if that 130 Farad bank means the batteries never see this ripple I will be happy and sleep a touch better.

Edit: this site is touchy when replying on my phone. 2nd part of message never posted for whatever reason

 

[wpns]

Good lord, a $1500 supercapacitor to try to solve a problem you aren’t even sure you have?

300 amps at 6.7 milliohms is 2V of ripple, so it may not even ‘solve’ your maybe-problem.

 

[CMDRZOD]

In any case I have no issue spending $1500 to help protect $35k of batteries. I need these to last many years to come and thats a small price to pay for their longevity.

I'd really like some more information on this topic to make a more informed decision. Unfortunately I've never seen this topic mentioned. I'd prefer to have a capacitor I didn't need than to have batteries fail prematurely.

 

[wpns]

In any case I have no issue spending $1500 to help protect $35k of batteries. I need these to last many years to come and thats a small price to pay for their longevity.

I'd really like some more information on this topic to make a more informed decision. Unfortunately I've never seen this topic mentioned. I'd prefer to have a capacitor I didn't need than to have batteries fail prematurely.

Fair enough, but a recent thread about microcycling it actually might improve battery life?

#WeDemandRigidlyDefinedAreasOfDoubtAndUncertainty

 

[740GLE]

I’d want to see it on a scope vs just a clamp meter.

That kind of current would heat up the cables even if it’s 4/0.

 

[CMDRZOD]

That's one thing I should've mentioned. I've got a JLN fuse holder with parallel 4/0 crimped and terminated on both sides of the holder.

I can feel warmth near the crimped ends.

 

[Vigo]

I hope you have patience for the skepticism.. it just seems like if it were a big issue it would already be widely discussed.

But then again i found the ac on both my PV and battery circuits 'the hard way' and then the PV issue did become heavily discussed here afterward so maybe it's real and just hasn't been widely 'discovered' yet?

On my battery circuit after getting shocked by AC on it i became pretty paranoid about what system-killing issue i might have, but after testing how much current was behind it with a light bulb and finding it to be pretty dang small, i simply grounded my battery bank and forgot about it and everything has been fine (at least, as far as i know) since then. That was probably a year or more ago.

 

[CMDRZOD]

Thats interesting Vigo, do you mean you were shocked when touching both 48v battery leads, or just the metal battery cases?

The only reason I decided to look into this is my fluke meter has a magnet and it vibrates like mad when near the battery leads, when there is seemingly just a small DC current going in/out of the batteries. When checking the DC amps all seems normal, but when measuring AC the meter reads very high, with high frequency.

Skepticism is a good thing. I welcome all data. Looks like my capacitor shipped and will be here early next week. I look forward to seeing the results once I connect it in parallel with the batteries.

 

[Solar Guppy]

AC ripple is normal and does not really have any impact on the battery cycle time as its is a faction of the voltage range of what is considered one cycle ( 100 to 0 SOC ) so there is no need to add parts

For added capacitors, you may not realize that your four Conext MPPT 100 have significant capacitor banks on the output already as they are design to work with lead-acid, which from a ripple perspective do less than Lifepo4 for ripple suppression

So spend that $$ and add batteries, the super capacitor will provided no improvements to your system.

Disclaimer: I was part of the team at Xantrex that designed the hardware you are using

 

[CMDRZOD]

AC ripple is normal and does not really have any impact on the battery cycle time as its is a faction of the voltage range of what is considered one cycle ( 100 to 0 SOC ) so there is no need to add parts

For added capacitors, you may not realize that your four Conext MPPT 100 have significant capacitor banks on the output already as they are design to work with lead-acid, which from a ripple perspective do less than Lifepo4 for ripple suppression

So spend that $$ and add batteries, the super capacitor will provided no improvements to your system.

Disclaimer: I was part of the team at Xantrex that designed the hardware you are using

I appreciate someone with intimate knowledge of my system chiming in. Still wondering about battery heating associated with that high current I see. Since the capacitor will arrive shortly I will update my findings. If the capacitor reduces battery heating by a few degrees I will call it a win.

 

[wpns]

Still wondering about battery heating associated with that high current I see.

Still wondering about the magnitude of the battery heating you see.

 

[CMDRZOD]

Here you go. This is yesterday.



Left axis Amps,volts
Right axis SOC, Temperature

 

[wpns]

Here you go. This is yesterday.

View attachment 201277

Left axis Amps,volts
Right axis SOC, Temperature

So 13 degrees C. Someone with more experience will have to tell you how normal that is. Interested to see how much it changes with the capacitor, and how much the capacitor heats up.

 

[740GLE]

It looks like temps climbs during charging which seems normal, during the AC ripple (high SOC) the temps don’t climb higher.

Is it because the temp is staying high, maybe a small blip, after SOC is high the concern?

If you saw say a 10c climb during the AC ripple after charging I’d be more concerned.

 

[CMDRZOD]

So 13 degrees C. Someone with more experience will have to tell you how normal that is. Interested to see how much it changes with the capacitor, and how much the capacitor heats up.

The cap comes with a temperature sensor output, just unsure of the exact type. I will be connecting this sensor to my automation once I have that figured out.

It looks like temps climbs during charging which seems normal, during the AC ripple (high SOC) the temps don’t climb higher.

Is it because the temp is staying high, maybe a small blip, after SOC is high the concern?

If you saw say a 10c climb during the AC ripple after charging I’d be more concerned.

My concern is the charge has nearly completed, but temperature has barely risen until my loads of opportunity ramp. The high load corresponds to the increased battery temperature. By the time the charge has reached 100A, the batt temp is stable.

The graph makes it very apparent that charging at 100A or below has very little effect on battery temperature. Once SOC and available PV watts are high, the loads start. That is when the temperature rise becomes very apparent to me.

Edit: I disagree with your last sentence. By the time the temperature really starts to rise, charging is nearly complete. I try to keep the battery at about 80% through load management in my automation.