Recomended battery voltage to charge with 20KW

fiblthip

New Member
Joined
Sep 26, 2021
Messages
9
Hi, IM relatively new to the solar space and this forum has already helped me greatly in understanding most of the important stuff. There are still some things I am not completely sure about yet, though.

What I can't seem to grasp is how to charge a single battery with high wattage of, for example, over 20KW. Would it be preferred to go with a 48V or 96V battery? At 48V, 20KW would be nearly 500A. Is there even a BMS that can handle that? If you go with a 96V battery, though, my questions is; where do you find 96V MPPT charge controller? And what if you want to double the power to 40KW? My guess is you would split up the solar array into multiple independent ones to charge independent smaller and lower voltage batteries, but then they might differ in voltage and couldn't be used together.

Sorry if this doesn't make any sense, but it is really confusing for me how people with large arrays deal with this.
 

FilterGuy

What, me worry?
Joined
Nov 26, 2019
Messages
4,448
Location
Los Gatos CA
Hi, IM relatively new to the solar space
how to charge a single battery with high wattage of, for example, over 20KW.
Wow! Nothing like jumping into the deep end to learn to swim!!! What are you wanting to power?
(BTW: I assume you mean 20KWh)

It is pretty rare to see someone go to 96V in the DIY space. 48V is about as high as anyone goes (at least on this forum). However, More and more of the power-wall products are going with high voltage batteries. I don't know much about them or the availability of compatible equipment. You might want to search for a power-wall forum and see if the DIY power-wall guys have any info.

Meanwhile, Where does the 500A come from??

Two banks of 16 280 Ah cells would be about 30KWh. ( 3.4V x 280Ah = 952Wh/Cell 952Wh/Cell x 32Cells = 30.464KWh.

Lets assume you want to charge those from empty to full in one sunny day and the insolation number for your area is 5 (5 equivalent hours of full sun). That means you need to run about 36464/5 = 6092.8 watts while charging. Assuming your system averages 54.4V, the current will be 6092/54.4 = 112A.
 

fiblthip

New Member
Joined
Sep 26, 2021
Messages
9
Thanks for the reply! Actually, I meant 20KW of charging power from Solar panels. Sorry, I could have expressed myself better.

What I mean to do is build an off grid system with a 55KWH battery and 25KW of solar panels. That's why I would need around 400A of charging current with a 48V system, as I will be getting maybe 20KW of maximum charging power at good conditions. The system needs to be so big because I want to run a 3.2KW appliance 24/7 (If you haven't guessed it, a bitcoin miner) and charge my electric car in the afternoon when the battery is full anyway.
 

Usangira

New Member
Joined
Dec 4, 2020
Messages
52
Thanks for the reply! Actually, I meant 20KW of charging power from Solar panels. Sorry, I could have expressed myself better.

What I mean to do is build an off grid system with a 55KWH battery and 25KW of solar panels. That's why I would need around 400A of charging current with a 48V system, as I will be getting maybe 20KW of maximum charging power at good conditions. The system needs to be so big because I want to run a 3.2KW appliance 24/7 (If you haven't guessed it, a bitcoin miner) and charge my electric car in the afternoon when the battery is full anyway.
Man,
What are your assumptions.
300% x 3.2KWH x 24 = 230kwhr per day assuming you won't be able to produce anything in next 2 days

230kwhrs divide by 5 sun hours .....

Interesting topic based on the consumption and criticality of the project.
 

fiblthip

New Member
Joined
Sep 26, 2021
Messages
9
Man,
What are your assumptions.
300% x 3.2KWH x 24 = 230kwhr per day assuming you won't be able to produce anything in next 2 days

230kwhrs divide by 5 sun hours .....

Interesting topic based on the consumption and criticality of the project.
Yes, that's right, I would be consuming 230KWH in 3 days. The thing is I live in a pretty sunny area and have relatively few cloudy days (40 days of rain last year). Also, the load is not strictly critical, It would still be OK to have no power for 10-20% of the year. I'm just trying to understand how one would try to build a system like this.
 

FilterGuy

What, me worry?
Joined
Nov 26, 2019
Messages
4,448
Location
Los Gatos CA
Thanks for the reply! Actually, I meant 20KW of charging power from Solar panels. Sorry, I could have expressed myself better.

What I mean to do is build an off grid system with a 55KWH battery and 25KW of solar panels. That's why I would need around 400A of charging current with a 48V system, as I will be getting maybe 20KW of maximum charging power at good conditions. The system needs to be so big because I want to run a 3.2KW appliance 24/7 (If you haven't guessed it, a bitcoin miner) and charge my electric car in the afternoon when the battery is full anyway.
Yup... a 2KW array is big and is going to generate around 400A on the battery side of a 48V system. I have never built anything that large.

I would investigate the high voltage systems I mentioned before....

If I were to build it on a 48V system at that scale, I think I would investigate breaking it down to several smaller systems. Each of the systems would use a 'stackable' inverter and tie them all together at the AC side.

1632758682644.png
The added cost is in the extra All-in-one units. The solar panel and battery costs are the same.
One nice thing about this approach is that it is easy to expand if you need to.

Since most of the all-in-one units are 2KW or larger, it might be overkill on the AC power. Consequently it would be worth looking at the cost of separate charge controllers and inverters.

1632759151800.png
 

fiblthip

New Member
Joined
Sep 26, 2021
Messages
9
Yup... a 2KW array is big and is going to generate around 400A on the battery side of a 48V system. I have never built anything that large.

I would investigate the high voltage systems I mentioned before....

If I were to build it on a 48V system at that scale, I think I would investigate breaking it down to several smaller systems. Each of the systems would use a 'stackable' inverter and tie them all together at the AC side.

View attachment 66540
The added cost is in the extra All-in-one units. The solar panel and battery costs are the same.
One nice thing about this approach is that it is easy to expand if you need to.

Since most of the all-in-one units are 2KW or larger, it might be overkill on the AC power. Consequently it would be worth looking at the cost of separate charge controllers and inverters.

View attachment 66542
Yes, that makes a lot of sense! In the separate systems, would it not be a problem if each battery pack had a slightly different SOC and voltage?
 

FilterGuy

What, me worry?
Joined
Nov 26, 2019
Messages
4,448
Location
Los Gatos CA
Yes, that makes a lot of sense! In the separate systems, would it not be a problem if each battery pack had a slightly different SOC and voltage?
I guess it is theoretically possible the systems could drift out of balance if they never had a chance to fully charge. However, once they are fully charged it kinda resets the balance.

EDIT: Added the following:

Any voltage differences in the battery banks would be compensated for by the inverter and it will put out the correct AC voltage. The only time the SOC differences would come into play is if one of the banks hit empty before the others. What would happen is that Inverter would drop out but the others would keep going. (This is actually an argument for having more aggrigate AC wattage than needed.... If one drops out, the others pick up the slack and the miner keeps digging for gold.)
 
Last edited:

fiblthip

New Member
Joined
Sep 26, 2021
Messages
9
I guess it is theoreticly possible the systems could drift out of balance if they never had a chance to fully charge. However, once they are fully charged it kinda resets the balance.
I would have to mount the independent solar arrays exactly the same, though, with the same tilt and azimuth, just for safety in case the batteries don't get a full charge for a few days. But yeah, it shouldn't be a problem as long as I make the arrays the same and have no shade over them. Thanks a lot for helping me understand this!

EDIT:
Read your edit, and yes, that is what I was worried about, one battery failing before the others. Now that you say it, of course the other inverters would just keep on working! All I have to do is oversize them to be able to handle the miner if one fails, which is probably a good idea anyway. As I only need 3KW it would not add much cost either.
 
Last edited:

MisterSandals

Participation Medalist
Joined
Nov 5, 2019
Messages
4,800
Location
Silicon Valley
I would have to mount the independent solar arrays exactly the same
Thinking outside the box a little, along the lines of 3 separate systems...

What if you broke your array into 3 different parts. 1/3 facing east-ish, 1/3 facing south, 1/3 facing west-ish?
This would get your system charging earlier in the morning (with 1/3 the power) and charging later in the evening (with 1/3 the power). You could charge 3x (on paper) as long at 1/3 the power.

Just a thought.
 

FilterGuy

What, me worry?
Joined
Nov 26, 2019
Messages
4,448
Location
Los Gatos CA
I would have to mount the independent solar arrays exactly the same, though, with the same tilt and azimuth, just for safety in case the batteries don't get a full charge for a few days. But yeah, it shouldn't be a problem as long as I make the arrays the same and have no shade over them. Thanks a lot for helping me understand this!
If you over-size the inverters, one can drop out and the system will keep humming along.

As an example if you had three 2KW inverters, the total AC capability would be 6KW. If one drops out you still have 4KW to run the 3.2KW minor.

The extreme of this is to have >3.2KW from each inverter, but I am not sure that would be worth it.


Warning: The following is not completely thought out... there may be holes in the idea.

I *think* that with proper layout, you could tie all the batteries together but never have one place that sees the full 400A of charge current.
This would keep the banks balanced.

1632763305367.png
This works conceptually, but careful thought is needed to ensure that there are not some corner cases where part of the system sees the full 400A. As an example, it should not happen, but if two of the banks were fully charged and the 3rd was half charged, all of the energy would try to flow to the 3rd battery. With them tied in parallel, this scenario should not happen. At the end of the charge cycle, the batteries will be accepting less current so even if they don't all hit full at the same moment, the current for the ones still charging should not be very high. My point is, all of these scenarios need to be thought through before you can be sure. If you can't be sure, then you are either going to have to put in breakers on the interconnects or wire everything big enough to handle the full 400A....which is exactly what we are trying to avoid.

EDIT: Added the following.
The more I think about the idea of tieing the batteries together, the more I realize that the interconnects need to be wired for the full 400A or there must be bi-directional breakers on the interconnects. During normal operation, the breakers would not pop, but if one of the inverters dropped out, the current flow would not be balanced and there could be higher current on the connections.
 
Last edited:

fiblthip

New Member
Joined
Sep 26, 2021
Messages
9
If you over-size the inverters, one can drop out and the system will keep humming along.

As an example if you had three 2KW inverters, the total AC capability would be 6KW. If one drops out you still have 4KW to run the 3.2KW minor.

The extreme of this is to have >3.2KW from each inverter, but I am not sure that would be worth it.


Warning: The following is not completely thought out... there may be holes in the idea.

I *think* that with proper layout, you could tie all the batteries together but never have one place that sees the full 400A of charge current.
This would keep the banks balanced.

View attachment 66548
This works conceptually, but careful thought is needed to ensure that there are not some corner cases where part of the system sees the full 400A. As an example, it should not happen, but if two of the banks were fully charged and the 3rd was half charged, all of the energy would try to flow to the 3rd battery. With them tied in parallel, this scenario should not happen. At the end of the charge cycle, the batteries will be accepting less current so even if they don't all hit full at the same moment, the current for the ones still charging should not be very high. My point is, all of these scenarios need to be thought through before you can be sure. If you can't be sure, then you are either going to have to put in breakers on the interconnects or wire everything big enough to handle the full 400A....which is exactly what we are trying to avoid.
Yes, of course I have to think everything through a lot with such a big system. And I don't feel safe relying solely on my imagination for possible dangerous scenarios, so I will definitely put breakers everywhere where a high current situation could happen and could be dangerous. Also, before building this, I will start with a smaller system just to gain some experience.
 

fiblthip

New Member
Joined
Sep 26, 2021
Messages
9
Thinking outside the box a little, along the lines of 3 separate systems...

What if you broke your array into 3 different parts. 1/3 facing east-ish, 1/3 facing south, 1/3 facing west-ish?
This would get your system charging earlier in the morning (with 1/3 the power) and charging later in the evening (with 1/3 the power). You could charge 3x (on paper) as long at 1/3 the power.

Just a thought.
Do you mean to break each independant array into three parts? That would be okay, and I will probably do that. All I want is every array to be exactly the same!
 

FilterGuy

What, me worry?
Joined
Nov 26, 2019
Messages
4,448
Location
Los Gatos CA

The more I think about the idea of tying the batteries together, the more I realize that the interconnects need to be wired for the full 400A or there must be bi-directional breakers on the interconnects. During normal operation, the breakers would [probably] not pop, but if one of the inverters dropped out, the current flow would not be balanced and there could be higher current on the connections.
 

fiblthip

New Member
Joined
Sep 26, 2021
Messages
9

The more I think about the idea of tying the batteries together, the more I realize that the interconnects need to be wired for the full 400A or there must be bi-directional breakers on the interconnects. During normal operation, the breakers would [probably] not pop, but if one of the inverters dropped out, the current flow would not be balanced and there could be higher current on the connections.
Why would the interconnects be at a higher current if an inverter died? The only high wattage is the charging wattage. If the inverter dies, that doesn't mean, the battery stops charging, assuming I have a separate inverter/CC system.
 

FilterGuy

What, me worry?
Joined
Nov 26, 2019
Messages
4,448
Location
Los Gatos CA
Why would the interconnects be at a higher current if an inverter died? The only high wattage is the charging wattage. If the inverter dies, that doesn't mean, the battery stops charging, assuming I have a separate inverter/CC system.
1632767999091.png
If everything is working as expected, I1, I2 and I3 will only be large enough to balance any small differences between the systems. From that point of view the interconnect wires can be quite small.

If, for whatever reason All-in-one 3 completely dropped out, then I3 would go up as the other two would try to charge or take power from the other two. I1 and I2 would also go up by about 1/2 of what I3 goes up. This means the wires have to be big enough to handle ~1/3 of the combined charge current of All-in-one 1 and all-in-one 2.

My point is this: I would be hesitant to assume I could think through all of the possibilities. Consequently, I would put bidirectional breakers on the interconnects to be sure that if something did happen, it would fail safely.

Note: I would size the interconnect fuses or breakers for ~125% of the charge current from one ALL-in-one and then size the wires to handle that current. That would cover all of the scenarios I can think of, but would fail safely if something I could not think of were to happen.

Warning: Some breakers are unidirectional. These would not work on the interconnects. They must be bidirectional.
 

fiblthip

New Member
Joined
Sep 26, 2021
Messages
9
View attachment 66566
If everything is working as expected, I1, I2 and I3 will only be large enough to balance any small differences between the systems. From that point of view the interconnect wires can be quite small.

If, for whatever reason All-in-one 3 completely dropped out, then I3 would go up as the other two would try to charge or take power from the other two. I1 and I2 would also go up by about 1/2 of what I3 goes up. This means the wires have to be big enough to handle ~1/3 of the combined charge current of All-in-one 1 and all-in-one 2.

My point is this: I would be hesitant to assume I could think through all of the possibilities. Consequently, I would put bidirectional breakers on the interconnects to be sure that if something did happen, it would fail safely.

Note: I would size the interconnect fuses or breakers for ~125% of the charge current from one ALL-in-one and then size the wires to handle that current. That would cover all of the scenarios I can think of, but would fail safely if something I could not think of were to happen.

Warning: Some breakers are unidirectional. These would not work on the interconnects. They must be bidirectional.
Yes, I see what you mean. I had thought that if I used a system with separate charge controller and inverter and not all in one, I would not have this problem, but now I realize only the charge controller would have to fail and it would have the same effect on the interconnects. I will 100% put breakers everywhere where high current situations can occur. I don't trust my imagination to pucture all possible dangerous scenarios!
 

mrzed001

Voice of reason
Joined
Apr 10, 2020
Messages
721
Location
Hungary - EU
View attachment 66566
If everything is working as expected, I1, I2 and I3 will only be large enough to balance any small differences between the systems. From that point of view the interconnect wires can be quite small.

If, for whatever reason All-in-one 3 completely dropped out, then I3 would go up as the other two would try to charge or take power from the other two. I1 and I2 would also go up by about 1/2 of what I3 goes up. This means the wires have to be big enough to handle ~1/3 of the combined charge current of All-in-one 1 and all-in-one 2.

My point is this: I would be hesitant to assume I could think through all of the possibilities. Consequently, I would put bidirectional breakers on the interconnects to be sure that if something did happen, it would fail safely.

Note: I would size the interconnect fuses or breakers for ~125% of the charge current from one ALL-in-one and then size the wires to handle that current. That would cover all of the scenarios I can think of, but would fail safely if something I could not think of were to happen.

Warning: Some breakers are unidirectional. These would not work on the interconnects. They must be bidirectional.


If I remember correctly most all-in-one inverters that can be paralleled require that the same battery is connected to each of them (and stop working if not).
So first parallel the 3 battery and the combined output to the 3 inverter.
The fuses for each battery string has to be able to take 0,5-1C of the battery (based on battery description what it can handle).
So if 1C 300A can be charged or discharged from every battery cell. Need a fuse at 300-320A for every battery string.
 
Last edited:
Top