Charge controller for large 1766AH, 48V battery bank with 6 hour peak sunshine

[Sunking]

Sadly (to me), if you just take 42000 * 1.5 you get the same 38 panel result. SMH.

Hate to say I told you so. There is no need to do all those calculations, It just introduces errors from to many calculations. 1,5 works everytime for MPPT, and 2 for PWM. For Pete's Sake John Wiles and 30 grad students took 3 years to develop the design process for Off-Grid. It is like the Wheel, you cannot improve the design. 1.5 will never leave you short. Any other method might come up short and leave you in the dark and destroyed batteries.

[Sunking]

I've been revising my loads/usage...here is what I will have and my tentative plan:

From 7AM till 6PM, 38 Panels will provide me with my 45162 Wh to be put back into my batteries.

Agreed or disagreed ?

Agree.

Personally I think you are out of your freaking mind. You are designing a system to fully discharge your batteries every day. They will not last more than a couple of months. With just one cloudy day you go dark and have to wait for a full day of sun to recharge before you can turn back on the power.

The battery required to provide you with 42 Kwh a day at 48 volts is 4200 AH, weighing some 13,000 pounds and cost you $30,000 every few years. That is just nucking futs.

With 5.23 Sun Hours and 42 Kwh daily usage [1.5 x 42000 wh / 5.23 Sun Hours will require:

Panel Wattage = 12,000 watts
MPPT Min Current = 12,000 wh/ 48 volts = 250 Amps
Battery AH Capacity @ 48 volts = 4200 AH weighing 13,000 pounds and cost you some $30,000 every 5 years.

You will be paying 70-cents per Kwh just in battery cost. The POCO sells it for 10 to 15 cents.

[sensij]

From 7AM till 6PM, 38 Panels will provide me with my 45162 Wh to be put back into my batteries

Agreed or disagreed ?

Sadly (to me), if you just take 42000 * 1.5 you get the same 38 panel result. SMH.

[SillySnakes]

I've been revising my loads/usage...here is what I will have and my tentative plan:

Solar Panels: 315W (@STC)
Power Consumption: 3500W
Hours of Use: 12 hours
Daily Wh: 42,000 Wh (42 kWh)

Inverter Eff: 93%

Daily Wh Adjusted: 42,000Wh/0.93 = 45162 Wh (45.162 kWh)

Daily Solar Panel Output (PWatts/SAM with carefully adjusted losses): 2.png



System Losses: 93% Solar Charger, 80% Battery Bank, 98% Wiring = 0.93*0.80*0.98=0.73

# of Solar Panels Needed: 45162 Wh/ (1649Wh)/(0.73) = 37.5 Panels = 38 Panels


From 7AM till 6PM, 38 Panels will provide me with my 45162 Wh to be put back into my batteries

Agreed or disagreed ?

[J.P.M.]

Sure if I am being paid to do it. Then I would take the time for all factors. But after doing several hundred designs, the difference is so small is insignificant.

With a grid tied makes no difference if you miss the mark over or under, it still works and every drop of power is used regardless. With a battery system there is only one side you can afford to miss on, the high side. Two designs have to be done. One for summer, and one for winter. You use the larger of the two which 90% of the time is winter. Come up short and you go dark, and destroy batteries, and/or use more generator fuel than necessary. If off-grid you had better have a generator to protect your battery investment.

So for me if I were being paid I run PV Watts, input a 1 Kw panel, 0% loss, orientation, and tilt. Click Go and see what the shortest days are. In the USA is going to be either December or January. So if 1 Kwh panel in December generates say 3200 watt hours. I know Sun hours is 3.2 is worse case. You cannot use yearly average as that is for Grid Tied only.

FWIW this is the method John Wiles and a few other pioneers came up with and tested at NMSU. At the time John and his grad students used a simple 100 watt panel with a MPPT Tracker and Dump Load to find Sun Hours. If that 100 watt panel generated 320 watt hours, Sun Hours = 3.2. That was well before PV Watts was ever conceived. John Wiles and his grad students then went on with JPL and NASA and developed PV Watts. No longer need to set out a panel for two months in winter. It is very simple and straight forward and a 5th grader can do it.

Yes on the Forum I multiply by 1.5 the daily watt hour usage. Works every time and never comes up short which is death. When I crunch the numbers once I know all the equipment, batteries, cable losses I have never ever came in lower than 1.3. For a DIY it is just not worth the time and effort to try to shave off 5 to 15% of panel power. Well unless you are talking something like a 2000 watt or more system. Otherwise 1.5 always works my friend, and 1.4 may fail my friend. Just KISS it.

Think of it this way JPM. You are the pilot of the plane. Your plane burns 13 gallons an hour. You calculate under current conditions will take 3 hours to get to your destination. As a pilot you want the plane to be light as possible and safe. So how much fuel do you have loaded, 3 x 13 = 39 gallons. Or do you follow the rules and make sure you have at least 52 gallons. DIY uses 39 gallons and crashes most of the time.

As my wife says, I don't have much of a butt. That is why I wear a belt and suspenders. No butt cracks or falling pants. .

I think I get the part about designing off grid systems for worst case Dec/Jan. No brainer once explained. If I got into off grid, I might do it different after I dug enough. Until then I'll defer to your judgment. Just sounded like you were applying the logic to all systems.

With not much of a butt, another good reason your an electron mongering EE and not a plumber. You'd never meet the plumber's crack requirement.

[sensij]

For a DIY it is just not worth the time and effort to try to shave off 5 to 15% of panel power. Well unless you are talking something like a 2000 watt or more system.

The OP is talking about a 14kW+ system. Worth the effort to do better analysis, and not just trust rules of thumb.

[Sunking]

Without getting into how off grid or battery system requirements (about which I know little) might impact, influence or change sizing or orientation considerations, I'd tend to stick with PVWatts for annual output estimates, and modify/iterate the array size and parameters to mesh with requirements of the after array equipment and application particulars. Maybe that's what your method does. But, respectfully, what you're suggesting seems to have more leaky variables in it than I'd use. I try to save my necessary fudge factors and their quantity until the end of the analysis. However, given my limited knowledge of off grid storage particulars, I don't have a better option to offer.

Sure if I am being paid to do it. Then I would take the time for all factors. But after doing several hundred designs, the difference is so small is insignificant.

With a grid tied makes no difference if you miss the mark over or under, it still works and every drop of power is used regardless. With a battery system there is only one side you can afford to miss on, the high side. Two designs have to be done. One for summer, and one for winter. You use the larger of the two which 90% of the time is winter. Come up short and you go dark, and destroy batteries, and/or use more generator fuel than necessary. If off-grid you had better have a generator to protect your battery investment.

So for me if I were being paid I run PV Watts, input a 1 Kw panel, 0% loss, orientation, and tilt. Click Go and see what the shortest days are. In the USA is going to be either December or January. So if 1 Kwh panel in December generates say 3200 watt hours. I know Sun hours is 3.2 is worse case. You cannot use yearly average as that is for Grid Tied only.

FWIW this is the method John Wiles and a few other pioneers came up with and tested at NMSU. At the time John and his grad students used a simple 100 watt panel with a MPPT Tracker and Dump Load to find Sun Hours. If that 100 watt panel generated 320 watt hours, Sun Hours = 3.2. That was well before PV Watts was ever conceived. John Wiles and his grad students then went on with JPL and NASA and developed PV Watts. No longer need to set out a panel for two months in winter. It is very simple and straight forward and a 5th grader can do it.

Yes on the Forum I multiply by 1.5 the daily watt hour usage. Works every time and never comes up short which is death. When I crunch the numbers once I know all the equipment, batteries, cable losses I have never ever came in lower than 1.3. For a DIY it is just not worth the time and effort to try to shave off 5 to 15% of panel power. Well unless you are talking something like a 2000 watt or more system. Otherwise 1.5 always works my friend, and 1.4 may fail my friend. Just KISS it.

Think of it this way JPM. You are the pilot of the plane. Your plane burns 13 gallons an hour. You calculate under current conditions will take 3 hours to get to your destination. As a pilot you want the plane to be light as possible and safe. So how much fuel do you have loaded, 3 x 13 = 39 gallons. Or do you follow the rules and make sure you have at least 52 gallons. DIY uses 39 gallons and crashes most of the time.

As my wife says, I don't have much of a butt. That is why I wear a belt and suspenders. No butt cracks or falling pants. .

[J.P.M.]

Agree and it is simple. Use a 1000 watt panel, 0% loss, input location orientation, and tilt will get you Sun Hours. Look for the month with lowest production. 1.5 Multiplier used takes care of the rest.

KISS

Without getting into how off grid or battery system requirements (about which I know little) might impact, influence or change sizing or orientation considerations, I'd tend to stick with PVWatts for annual output estimates, and modify/iterate the array size and parameters to mesh with requirements of the after array equipment and application particulars. Maybe that's what your method does. But, respectfully, what you're suggesting seems to have more leaky variables in it than I'd use. I try to save my necessary fudge factors and their quantity until the end of the analysis. However, given my limited knowledge of off grid storage particulars, I don't have a better option to offer.

[Sunking]

Provided the user knows how to use PVWatts, or particularly SAM, get the inputs reasonably close to reality and applies some judgment to the results. GIGO.

Agree and it is simple. Use a 1000 watt panel, 0% loss, input location orientation, and tilt will get you Sun Hours. Look for the month with lowest production. 1.5 Multiplier used takes care of the rest.

KISS

[J.P.M.]

Look at his chart for that day. It tells him he receives 5.2 Sun Hours. No reason to make it anymore complicated. 1649 wh / 315 w = 5.23 Sun Hours. You use PV watts to find th eworse case month and use the Sun Hours. Works everytime.

Provided the user knows how to use PVWatts, or particularly SAM, get the inputs reasonably close to reality and applies some judgment to the results. GIGO.

[Sunking]

Read what BackwoodsEE For a system of this size, I would ignore sunking's rules of thumb for solar newbies and use better engineering tools like PVWatts or SAM to estimate solar output.

Look at his chart for that day. It tells him he receives 5.2 Sun Hours. No reason to make it anymore complicated. 1649 wh / 315 w = 5.23 Sun Hours. You use PV watts to find th eworse case month and use the Sun Hours. Works everytime.

[littleharbor]

Another thing to keep in mind is that only the bulk stage of charging uses all available panel power, but that only gets you 80% charged or so. Once the controller moves into absorb stage at a fixed voltage, there current the battery can accept becomes limited. This means not all of the energy available from the array is used, and is effectively another loss factor to consider when sizing the array based on daily energy of Ah output.

And a good reason to use opportunity loads if you can. Many off grid folks will delay laundry, vacuuming or other high load activities till afternoon when this otherwise wasted potential power is available.

[sensij]

Alright, I think I understand now. I read the thread Sunking pointed to.

So here is a table of my future panel output wattage as a function of time (thanks to PWatts/SAM simulation). The panel is a 315W STC panel. So in calculating how much total AH this panel will produce a day, I divide each Watts by 48V during each hour....and then sum up all the Amperage per hour to give me total AH of capacity generated at the end of the day, correct ? From this I can plan out how many panels/controllers I need to recharge my 1160AH bank properly per day

Just keep in mind that the power estimate produced by PVWatts or SAM will have assumptions on losses, and the default loss values are for grid tie systems, not battery systems. It would be appropriate to use the DC power modeled by each system as a starting point to which charge controller and battery losses would be applied.

Another thing to keep in mind is that only the bulk stage of charging uses all available panel power, but that only gets you 80% charged or so. Once the controller moves into absorb stage at a fixed voltage, there current the battery can accept becomes limited. This means not all of the energy available from the array is used, and is effectively another loss factor to consider when sizing the array based on daily energy of Ah output.

[sensij]

I agree that a high end MPPT CC will be able to do a lot more then a cheap MPPT. But most 60Amp MPPT CC's will limit the output to no more than 1.10% of that 60amp rating and some max out at 100% of the rating depending on both the internal components and software.

Read what BackwoodsEE wrote a couple more times. He is not saying that the controller produces more than it's rating. He is explaining how the operating point of the array is adjusted to control the power. This is very basic mppt functionality... Cheap ebay junk might not have it, but most legit controllers are capable of controlling array power in this way, certainly the midnight controllers the op is telling about. They can handle arrays much larger than you are suggesting, with no extra heat dissipation at the controller.

For a system of this size, I would ignore sunking's rules of thumb for solar newbies and use better engineering tools like PVWatts or SAM to estimate solar output. The table the OP posted looks like the right approach.

Back to one of the original questions, cables and busbars are sold that are rated for these high currents. It is important that all terminations are made in a way that meets the rating... Hydraulic crimps, bolts torqued to the proper value, etc. Don't cut corners on here.

[SunEagle]

I've watched my TS-MPPT-60 in action with a real-time voltage-current tracing device I designed and built. What the charge controller does to limit battery current is pretty slick, and does not dissipate heat anywhere downstream of the panels. It simply adjusts where it operates on the PV array's IV curve so that less power is drawn from the array.

For example, right now my toy 3-panel array is producing a mere 110-120W to float charge the battery and run a few isolated loads. The PV current is hovering around 98-99V, well above the maximum power point. If I turned on more load (I've done this many times with big incandescent lights while watching the real-time plot), it would slide the voltage down to the MPPT and the PV current would rise dramatically. Occasionally, I've also seen it move the PV voltage up, toward a MPPT that was higher where it got left the last time it reduced power draw.

It's provided me hours of free geek entertainment watching the current voltage and current point slide left/up and right/down along the IV curve as the charge controller adjusts its production to match the inverter's consumption on a float-charging battery. A good MPPT charge controller is a sophisticated device indeed.

I agree that a high end MPPT CC will be able to do a lot more then a cheap MPPT. But most 60Amp MPPT CC's will limit the output to no more than 1.10% of that 60amp rating and some max out at 100% of the rating depending on both the internal components and software.