Tweet
From what I can see, you have two distinct problems.
1. What to do with the existing panels to make a perfect system from scratch.
2. What to do with what you have inherited in the meantime.
Unlike Butch's concept (posts #2 and #6) that you utilize a single battery bank I propose the opposite. I suggest that you split the system up into 'pods' that are completely independent to supply smart inverters that are AC coupled on the output. This appears to be the original 'concept', two pods.
I have simplified the analysis by presuming that each pod is nominally 48V. There is no reason they couldn't be 24V, at least with extremely small pods, but for maintenance purposes you want to keep a 'commonality' of equipment (inverters) and the larger pods demand the 48V and there is no good reason for anything lower. Also, I have limited discussion to FLA batteries. I hear you begging for Lithium, but sometimes if you want to farm you have to have one person to step up to learn about lead. I would rather they learn FLA than me guess at the right charge amperage. The last presumption is that you have 60 290W panels with three spares.
The following table makes some suggestions as to pod size to utilize all 60 panels. (now I see why I haven't seen the table tool used much in posts)
Concerning the 'what to do with what I got scenario': [presented in a 'thinking out loud' approach]
Considering that you have two 3800 W inverters, I suggest that we look at the 15 panel pods at 4350 Watts. 3800/4350 = 0.81 which seems like a good match. It will provide 4350/48 = 90 amps max charging current. To use your 225Ah batteries we are looking at 4 parallel strings to make that equal to a C/10 charging current. We do not want to go to four parallel battery strings due to balancing issues, plus our charge controller only puts out 80A, so lets look at the 3480W pod.
This will provide 3480W/48V = 72 max charging amps. Three strings of 225 Ah = 675 Ah. So that would be 675 Ah charged by 72 amps = about a C/9.4 charge rate so we're right in the ballpark. This is very attractive.
Since we have panel limited our current we can go oversized on the charge controller/inverter if we desire.
But WAIT. We only have 16 of the 6V 225 Ah batteries to work with. In other words only two strings of 48V which would come out to 450 Ah. A C/10 charge current on that bank is 45 amps. 45A x 48V = 2160W. We don't have a 2160W pod, the 2900W pod gives us 2900/48 = 60A which is a 450 Ah / 60 = C/7.5 charge rate, a touch high. If we used 8 panels we would have 1800W, that works out to 1800/48 = 40 amps. 450 Ah / 40A = a C/11.25 charge rate, a bit wimpy but perhaps a better value for the aged and neglected battery bank.
Whenever we design to a current limit, we have to revisit charging and load. I assume we do not wish to exceed 50% depth of discharge which would be 450 Ah x 48V x 0.5 = 10.8 kWh. The daily harvest with eight panels is 1800W x 5 sun hours = 9kWh per day. Drawing 9kWh per day will be fine as far as the batteries go. Note that drawing it all daily will make it difficult to recharge on solar only after a cloudy day. Up to you how much less you wish to use (or how you'll augment the solar). With the high charge option, harvest is 2900W x 5 sun hours = 14.5kWh, much more than enough to budget the use of the full 10.8 kWh depth of discharge proviso.
So, there you have it. For the moment connect all 16 of your batteries into two strings of 8 in series for 48V 450 Ah. Since we panel limited the current, we can go with one of the current charge controllers and one of the current inverters. Your choice of whether to hook up 8, 9, or 10 panels will be decided by your charging current preference, the condition of the batteries, and the charge controller hookup schema. By the way, if your charge controller does have a 150 V input limit, and your nominal panel Voc is 45 volts, I think I would stick to only two of these panels in series which probably removes the 9 panel option. On the other hand, if temperatures absolutely won't go too low, the nine panel option might be very attractive charging-wise with three strings of three panels in series.
As far as the future, building around the panels from scratch, you need to decide if you want off grid inverters (with chargers) or separate charge controllers. Regardless, the inverter choice should drive the whole system design and the pods to support them built to match. You can see how which battery type/size comes into play in the future is extremely important and expensive and should be a primary consideration in inverter selection. The system pod design should NOT rely on panel wattage to limit charging current (which is what I did to re-use your controllers and inverter) - I have always hated that approach.
Please note that the balanced pods is just one approach, if you run into a equipment blend that suggests two 20 panel pods and two 10 panel pods, or any other combination (even 9 panel), you should go for it.
You asked more things about loads and cloudy days and 'tapping' battery strings, etc. but I wish to see how badly this gets blown out of the water (and to what extent you understand it) before I address that. Last thing I want to do is kill any fish.
I look forward to following any inverter and battery discussion for a "perfect from scratch" solution.
Edit: Ugh, I just realized that his current inverters probably won't handle 48V. I'll add another post addressing what he can do with two 24V inverters if nobody else does. I need something to give me about 24 kV to the groin area every time I type the word 'inverter'.
1. What to do with the existing panels to make a perfect system from scratch.
2. What to do with what you have inherited in the meantime.
Unlike Butch's concept (posts #2 and #6) that you utilize a single battery bank I propose the opposite. I suggest that you split the system up into 'pods' that are completely independent to supply smart inverters that are AC coupled on the output. This appears to be the original 'concept', two pods.
I have simplified the analysis by presuming that each pod is nominally 48V. There is no reason they couldn't be 24V, at least with extremely small pods, but for maintenance purposes you want to keep a 'commonality' of equipment (inverters) and the larger pods demand the 48V and there is no good reason for anything lower. Also, I have limited discussion to FLA batteries. I hear you begging for Lithium, but sometimes if you want to farm you have to have one person to step up to learn about lead. I would rather they learn FLA than me guess at the right charge amperage. The last presumption is that you have 60 290W panels with three spares.
The following table makes some suggestions as to pod size to utilize all 60 panels. (now I see why I haven't seen the table tool used much in posts)
| Panels/Pod | Pod Watts | C/10 Amps |
| 60 | 17,400 | 363 |
| 30 | 8700 | 181 |
| 20 | 5800 | 121 |
| 15 | 4350 | 91 |
| 12 | 3480 | 72 |
| 10 | 2900 | 60 |
| 6 | 1740 | 36 |
Concerning the 'what to do with what I got scenario': [presented in a 'thinking out loud' approach]
Considering that you have two 3800 W inverters, I suggest that we look at the 15 panel pods at 4350 Watts. 3800/4350 = 0.81 which seems like a good match. It will provide 4350/48 = 90 amps max charging current. To use your 225Ah batteries we are looking at 4 parallel strings to make that equal to a C/10 charging current. We do not want to go to four parallel battery strings due to balancing issues, plus our charge controller only puts out 80A, so lets look at the 3480W pod.
This will provide 3480W/48V = 72 max charging amps. Three strings of 225 Ah = 675 Ah. So that would be 675 Ah charged by 72 amps = about a C/9.4 charge rate so we're right in the ballpark. This is very attractive.
Since we have panel limited our current we can go oversized on the charge controller/inverter if we desire.
But WAIT. We only have 16 of the 6V 225 Ah batteries to work with. In other words only two strings of 48V which would come out to 450 Ah. A C/10 charge current on that bank is 45 amps. 45A x 48V = 2160W. We don't have a 2160W pod, the 2900W pod gives us 2900/48 = 60A which is a 450 Ah / 60 = C/7.5 charge rate, a touch high. If we used 8 panels we would have 1800W, that works out to 1800/48 = 40 amps. 450 Ah / 40A = a C/11.25 charge rate, a bit wimpy but perhaps a better value for the aged and neglected battery bank.
Whenever we design to a current limit, we have to revisit charging and load. I assume we do not wish to exceed 50% depth of discharge which would be 450 Ah x 48V x 0.5 = 10.8 kWh. The daily harvest with eight panels is 1800W x 5 sun hours = 9kWh per day. Drawing 9kWh per day will be fine as far as the batteries go. Note that drawing it all daily will make it difficult to recharge on solar only after a cloudy day. Up to you how much less you wish to use (or how you'll augment the solar). With the high charge option, harvest is 2900W x 5 sun hours = 14.5kWh, much more than enough to budget the use of the full 10.8 kWh depth of discharge proviso.
So, there you have it. For the moment connect all 16 of your batteries into two strings of 8 in series for 48V 450 Ah. Since we panel limited the current, we can go with one of the current charge controllers and one of the current inverters. Your choice of whether to hook up 8, 9, or 10 panels will be decided by your charging current preference, the condition of the batteries, and the charge controller hookup schema. By the way, if your charge controller does have a 150 V input limit, and your nominal panel Voc is 45 volts, I think I would stick to only two of these panels in series which probably removes the 9 panel option. On the other hand, if temperatures absolutely won't go too low, the nine panel option might be very attractive charging-wise with three strings of three panels in series.
As far as the future, building around the panels from scratch, you need to decide if you want off grid inverters (with chargers) or separate charge controllers. Regardless, the inverter choice should drive the whole system design and the pods to support them built to match. You can see how which battery type/size comes into play in the future is extremely important and expensive and should be a primary consideration in inverter selection. The system pod design should NOT rely on panel wattage to limit charging current (which is what I did to re-use your controllers and inverter) - I have always hated that approach.
Please note that the balanced pods is just one approach, if you run into a equipment blend that suggests two 20 panel pods and two 10 panel pods, or any other combination (even 9 panel), you should go for it.
You asked more things about loads and cloudy days and 'tapping' battery strings, etc. but I wish to see how badly this gets blown out of the water (and to what extent you understand it) before I address that. Last thing I want to do is kill any fish.
I look forward to following any inverter and battery discussion for a "perfect from scratch" solution.
Edit: Ugh, I just realized that his current inverters probably won't handle 48V. I'll add another post addressing what he can do with two 24V inverters if nobody else does. I need something to give me about 24 kV to the groin area every time I type the word 'inverter'.
Comment