More PV versus more Batts?

[mesquito_creek]

I am looking to add some insurance to my system and need advise on where to stretch the capabilities. I am running 4000 watts of pv, 21,000 watt hours of batts and a mppsolar LV6048 to power an RV off grid. RV has two Colman 15000 btu AC units, residential fridge and lots of vampires. Minus the AC load I think I over killed the system and for winter and spring I hardly ever used more than maybe 20% of the batteries on any given 24 hour cycle.

with summer time here I need to run AC overnight. During the day from 10-4 I can run 2 AC and the base load off of PV for the most part and head into the evening with my batts topped off.

I am sitting here at 8 am and the sun is so far to the north it’s almost hitting the back of my solar array and I am wondering if I make sense to add more panel pointing west and east and just extend my solar production. I thought is if I can produce more watts sooner in the day and longer into the evening that is more economical than adding more batteries. My mppt charger has an open input for another 4000 watts not being used. I would get some used panels and point 2000 watts west and 2000 watts east? Is this a good strategy or would the experts say always just add more panels facing south.

My weak spot is I don’t have 2 or 3 full days of storage under AC loads. I am in southern Utah 37 degrees and I am willing to go short on storage because the sun is nearly always available. If it is cloudy then my AC requires way less use and I might have plenty of battery.

 

[snoobler]

How much surplus solar do you have now?

Do you really need to run both A/C units at night? My son is living in a Class A with dual A/C here in Mesa, AZ. We've had several days in a row of "EXCESSIVE HEAT WARNING REMAINS IN EFFECT...". He's been comfortable with a single A/C unit running, especially at night.

It's a viable strategy, but it's important to know what you're getting. Use PVWatts in link #6 in my signature to simulate what a theoretical array would produce for you. You can download hourly data to see how it benefits morning/evening.

I find it easiest to model the arrays separately and combine them in a spreadsheet. On that same site is the SAM software. It allows you to model an entire array, but I've never invested the time to figure out how to use it.

 

[MichaelK]

Similar to my own solution. My main arrays are all South facing, but are on rotating mounts. For more power at 8am, I can rotate the panels facing
East, then rotate West in the afternoon. I can guaranty 2000+W from 8am till 4pm for running my well-pump.

After about 4:30 though my primary arrays start getting shaded no matter how westward I turn them. So, I installed a 5th 1000W array facing due West in a separate location. So, just as my primary arrays are dropping off, the westward array is at full output.

 

[noenegdod]

My weak spot is I don’t have 2 or 3 full days of storage under AC loads. I am in southern Utah 37 degrees and I am willing to go short on storage because the sun is nearly always available. If it is cloudy then my AC requires way less use and I might have plenty of battery.

It sounds like you dont have a weak spot. It seems as though your thought process has a disconnect.

You dont have 2 or 3 full days of storage under the AC loads you would have when its sunny, but when its not sunny and you cant produce enough power to satisfy the sunny day loads, you AC load is actually way less, so you don't need to produce as much power. It sounds like your exactly where you would want to be, unless of course you are looking for someone to provide you with an excuse to buy more batteries or panels.

 

[noenegdod]

Holy crap, it appears I have stuttered while I was clicking post reply. I cant seem to figure out how to delete posts. Can someone either delete those or tell me how to.

 

[mesquito_creek]

@snoobler I currently agree that I only need one AC at night. BUT that does force my wife to sleep in the same room with me! This is a seasonal place for us so we are committed to small inconvenience like a single AC. We only need AC for a couple months since we are Phoenix people! When I woke up this morning after running a single AC my battery bank was about 52.4v and the AC had not turned on for several hours. But I had morning temps of 60 degrees. We are headed into a heat wave this week with morning lows of 75, so I will get a real time test if I stick around for it.

I started making about 1000 watts of excess PV at 8 am so I am probably over thinking things. Investing in a RV shade structure probably would be more effective even though the cost will be many times more..

 

[mesquito_creek]

@snoobler I really must find some good training/learning material on what state of charge means because I am ignorant! I have a lifepo4 4v battery bank of 4 BB Hsky 5.3kw batteries. When I charge them up to 57v and put a 2-3K watt load on them they quickly discharge down to ~53v. I learned in another thread about how that happens because of the "flat part of the curve". So I think I mostly understand that aspect.

So I am running my full load yesterday, which is 2 ACs a fridge and other stuff like TVs etc... That load is about 3000 watts and my PV array is mostly keeping up, but toward 5 pm or so I am at a slight deficiency, so my state of charge is trending down ward. No biggy, but I have about 3 hours before the peak heat, no more solar production and the sun goes down to lessen the AC loads. My victronconnect is showing that I am using about 1500-2500 watts off and on as we go into the evening.

We go to bed around 10 pm and I shut off the main AC and leave the bedroom AC running in "constant fan mode". My battery bank shows approx high 51.x v (51.8 for example). The state of charge trend line on my victron is continuing to trend downward. I wake up at about 1 am and check the voltage and I am now ~50v. The low end of 50v, for example 50.2v. I turned off the AC from constant fan to regular mode and it cycles on/off 2 more quick cycles and then never turns back on because the outside temps are in the 60s.

I wake up this morning at 6 am and the battery bank shows 52.2v? I am now once again showing how ignorant I am. I first off was expecting to see my voltage in the 49 range or at least not higher than when I went back to bed. I am looking at the "curve" graph and have no idea if I was toward the end of it.

Will the state of charge at the end of the curve act like the start of the curve where I will get to something like 49.5 and then quickly discharge down to my setting of 46.5 low dc cut off? Did I have plenty of battery left and I am just being too conservative etc.. ? and finally, what is going on with the bounce back to 52.2v in the middle of the night? What is my state of charge at 52.2v, I am looking at all sorts of different charts with different values and in general I think anything in the 52.x range is a decent state of charge ie. 70% or so?

 

[snoobler]

@snoobler I really must find some good training/learning material on what state of charge means because I am ignorant! I have a lifepo4 4v battery bank of 4 BB Hsky 5.3kw batteries. When I charge them up to 57v and put a 2-3K watt load on them they quickly discharge down to ~53v. I learned in another thread about how that happens because of the "flat part of the curve". So I think I mostly understand that aspect.

Voltage rises with charge and drops with discharge. When those currents are removed/lessened, the voltage "rebounds" in the opposite direction.

If you cut charge and left that battery sitting with no loads, you would see it drop over time to 54-55V yet still retain near 100% of its charge.

The relationship between SoC and voltage is weak.

So I am running my full load yesterday, which is 2 ACs a fridge and other stuff like TVs etc... That load is about 3000 watts and my PV array is mostly keeping up, but toward 5 pm or so I am at a slight deficiency, so my state of charge is trending down ward. No biggy, but I have about 3 hours before the peak heat, no more solar production and the sun goes down to lessen the AC loads. My victronconnect is showing that I am using about 1500-2500 watts off and on as we go into the evening.

We go to bed around 10 pm and I shut off the main AC and leave the bedroom AC running in "constant fan mode". My battery bank shows approx high 51.x v (51.8 for example). The state of charge trend line on my victron is continuing to trend downward. I wake up at about 1 am and check the voltage and I am now ~50v. The low end of 50v, for example 50.2v. I turned off the AC from constant fan to regular mode and it cycles on/off 2 more quick cycles and then never turns back on because the outside temps are in the 60s.

I wake up this morning at 6 am and the battery bank shows 52.2v? I am now once again showing how ignorant I am. I first off was expecting to see my voltage in the 49 range or at least not higher than when I went back to bed. I am looking at the "curve" graph and have no idea if I was toward the end of it.

That's what I described above - the voltage has risen due to reduction of loads. You've gone from 2000-3000W to maybe just 200W. HUGE difference.

Need to think of the flat part of the voltage curve in context of current. For a given load the voltage curve is very flat, but if you increase/decrease current, you lower/raise the entire curve. This chart illustrates that:



Given the size of your battery, I doubt you're ever pulling more than 1C, so you're most interested in the top 3 lines.

Will the state of charge at the end of the curve act like the start of the curve where I will get to something like 49.5 and then quickly discharge down to my setting of 46.5 low dc cut off? Did I have plenty of battery left and I am just being too conservative etc.. ? and finally, what is going on with the bounce back to 52.2v in the middle of the night? What is my state of charge at 52.2v, I am looking at all sorts of different charts with different values and in general I think anything in the 52.x range is a decent state of charge ie. 70% or so?

Again, state of charge is mostly independent of the voltage with only loose correlation. The SoC is calculated based on the Ah used vs. the Ah available.

Do you have a battery monitor that actually reports % state of charge?

 

[mesquito_creek]

Do you have a battery monitor that actually reports % state of charge?

I have the VictonConnect Shunt that Will uses in the videos. And to your point, its state of charge reading vs the voltage is quite polar opposite at times.

My battery BMSs have bluetooth bms but no support or app yet. But I can read something off of them with an unsupported Chinese app I downloaded.

Here are some current screen shots, I am running an AC unit at the moment and have started making some PV power about an hour ago.

 

[snoobler]

Once it's properly configured, trust the Smartshunt. I have a BMV-702.

What is your solar charge controller's absorption voltage?

 

[mesquito_creek]

Once it's properly configured, trust the Smartshunt. I have a BMV-702.

What is your solar charge controller's absorption voltage?

56.5v

 

[snoobler]

Set your charged voltage to 56.3V

Set your tail current to 5%

For solar charging, Victron recommends you base 100% SoC criteria on absorption voltage (-0.2V) instead of float.

If your float is above 54.4V, I would lower it to 54.4V.

You may notice the Shunt SoC behavior changes, but it is likely a little more accurate than before.

 

[mesquito_creek]

If your float is above 54.4V, I would lower it to 54.4V.

You are instructing me to set my MPPT controllers float to 54.4 correct? ...thank you for all the help!

 

[snoobler]

Yes.

Summary:

Smartshunt charged to 56.3V
Smartshunt tail current to 5%
MPPT charge controller float to 54.4V

 

[Hedges]

"More PV versus more Batts?"

I favor more PV where there is a choice. (of course only batteries can provide power without sun.)
PV panels can have a useful life of 25 to 50 years. Degradation rates vary; I want panels likely to last and continue producing well.
Cost of panels is around $0.15 to $0.50/W. With 5.5 hours average effective sun, over 10 years that amortizes out to $0.007 to $0.025/kWh. Declining further over 20 year or longer use, but time-value of money so I use 10 years.

Additional cost for charge controller, although over-paneling so the extra PV only helps on cloudy days, and multiple orientations for more hours, help reduce SCC cost. With single PV array orientation, a Midnight Classic 150 would cost $0.0065/kWh over 10 years.

Batteries such as AGM or commercial lithium cost around $0.50/kWh (lithium is coming down over time), and DIY LiFePO4 looks like about $0.039/kWh. But only if it lasts the equivalent of 3500, 100% DoD cycles. I recently read a report on testing of dozens of brands of commercial batteries. A couple delivered 3500 cycles, but 95% or more failed, requiring repair or replacement.

So to me, PV & SCC costs 1/3 to 1/2 what batteries do and is more reliable. Therefore if over-building PV 2x or 3x lets me get by without much battery, that's what I prefer. My grid-backup system has 10kW (AC) from PV and 14 kWh (usable) of AGM. When grid is down I run A/C while the sun shines but not when PV output drops low in the evening. Since I don't live off-grid, if I guessed wrong on system sizing I won't suffer the consequences for long.

 

[rin67630]

It sounds like you dont have a weak spot. It seems as though your thought process has a disconnect.

You dont have 2 or 3 full days of storage under the AC loads you would have when its sunny, but when its not sunny and you cant produce enough power to satisfy the sunny day loads, you AC load is actually way less, so you don't need to produce as much power. It sounds like your exactly where you would want to be, unless of course you are looking for someone to provide you with an excuse to buy more batteries or panels.

If you know your figures: panels, orientation, tilt, and your consumption you may use this tool to make projections with various battery capacities and check how much energy is unused, respectively missed and how frequent you end up with an empty battery:
https://re.jrc.ec.europa.eu/pvg_tools/en/tools.html#SA

 

[snoobler]

If you know your figures: panels, orientation, tilt, and your consumption you may use this tool to make projections with various battery capacities and check how much energy is unused, respectively missed and how frequent you end up with an empty battery:
https://re.jrc.ec.europa.eu/pvg_tools/en/tools.html#SA

Thank you for that. I NOW remember you posting that before, and I was looking for it earlier.

I dug into SAM and kinda figured out how to represent 3 subarrays (East, South, West) using the same type of panel, but it kinda sucks to try and approximate off-grid stuff. Off the top of your head, will the tool you linked permit modeling of multiple arrays with different panels?

 

[mesquito_creek]

"More PV versus more Batts?"

I favor more PV where there is a choice. (of course only batteries can provide power without sun.)
PV panels can have a useful life of 25 to 50 years. Degradation rates vary; I want panels likely to last and continue producing well.
Cost of panels is around $0.15 to $0.50/W. With 5.5 hours average effective sun, over 10 years that amortizes out to $0.007 to $0.025/kWh. Declining further over 20 year or longer use, but time-value of money so I use 10 years.

Additional cost for charge controller, although over-paneling so the extra PV only helps on cloudy days, and multiple orientations for more hours, help reduce SCC cost. With single PV array orientation, a Midnight Classic 150 would cost $0.0065/kWh over 10 years.

Batteries such as AGM or commercial lithium cost around $0.50/kWh (lithium is coming down over time), and DIY LiFePO4 looks like about $0.039/kWh. But only if it lasts the equivalent of 3500, 100% DoD cycles. I recently read a report on testing of dozens of brands of commercial batteries. A couple delivered 3500 cycles, but 95% or more failed, requiring repair or replacement.

So to me, PV & SCC costs 1/3 to 1/2 what batteries do and is more reliable. Therefore if over-building PV 2x or 3x lets me get by without much battery, that's what I prefer. My grid-backup system has 10kW (AC) from PV and 14 kWh (usable) of AGM. When grid is down I run A/C while the sun shines but not when PV output drops low in the evening. Since I don't live off-grid, if I guessed wrong on system sizing I won't suffer the consequences for long.

Thanks for the input. My hybrid controller (mppsolar lv6048) has an additional charge controller that I am not using, so this helps makes the case for a targeted PV array to the west. With some mechanical improvements (shade structure) for the RV off grid I think I may be able to avoid more batteries... OR just avoid this place altogether when its too hot!