Before I make ANOTHER mistake

South Africa

Before I commit on these two system:

For both systems:
- 2 pole fuses for all incoming strings, as well as between batteries and controller with the correct amp fuses per holder.
- 1 correctly specced fuse on inverters positive line,
- 5.5 hours daylight, Cape Town, South Africa.
- For boths systems grid is the backup for rainy days, extended winter conditions which means in summer, add more loads, winter remove them.
Cabling: as per requirements per system.


Note 1: BEFORE the load is started during daylight hours, the batteries will be fully charged.
Note 2: Batteries are for lights at night, and to connect the equipment. :-)
Note 3: Daytime use is limited to 10% DOD for 2 reasons:
1) To have enough power left for lights at night.
2) IF there is a power failure, on System 2 only, and there is no solar, then have spare capacity as:
2a.1) UPS for computer equipment, with limited lights,
2a.2) or lights and no computers.



System one:
- Primary function is lights, secondary are daytime loads: 1190WH per night.
- 2 x 200w Tenesol Panels, connected in series.
- MS MPPT 15 controller.
- 24v 102ah batteries.
- Victron 350w pure sinewave inverter.

Questions for system 1:
1) What is the battery bank A/H I must / can put on based on panels?
2) For lead acid deep cycle, are the 102ah the best ah ones to use, or go smaller ah batteries?
3) After batteries are charged, what spare power can I anticipate per day?



System two:
- Primary function is daytime loads, secondary is lights. Note: Less batteries = faster charging = quicker to get to powering equipment from solar. :-)
- Load 1: At night, 1050WH of lights.
- Load 2: Once batteries are fully charged, I want to power as many devices as the panels (maxed) and inverter (comfortably) can handle.
- Panels, 6 x 310w = 1860w, connected series / parallel.
- Controller Outback Flexmax 60 or 80
- 24v 102ah batteries.
- Victron 1600w inverter.

Questions for system 2:
1) What is the battery bank A/H I must / can put on based on panels?
2) For lead acid deep cycle, are the 102ah the best ah ones to use?
3) After batteries are charged, what spare power can I anticipate per day?

bill von novak

Re: Before I make ANOTHER mistake

South Africa wrote: »

1) What is the battery bank A/H I must / can put on based on panels?

Probably don't want less than 500 amp-hours (50-60 amp charge current)

2) For lead acid deep cycle, are the 102ah the best ah ones to use?

If you want to go cheap, go with GC2's. They are commonly available, reasonably reliable and fairly cheap.

3) After batteries are charged, what spare power can I anticipate per day?

About 1200 watts during peak daylight hours per your numbers.

Cariboocoot

Re: Before I make ANOTHER mistake

bill von novak wrote: »

Probably don't want less than 500 amp-hours (50-60 amp charge current)

For 400 Watts of panel @ 24 Volts? I think you got this confused with the other system. About 13 Amps and 102 Amp hours is fine. 204 is a bit heavy on batteries. This is the battery size he has to work with.

This is not going to supply 1190 Watt hours at night, however. In fact 400 Watts of panel running 'full battery' (all power in and out) will manage about 1000 Watt hours total with 5 hours equivalent good sun. Putting in more battery won't change that. 102 Amp hours is only going to manage about half the desired nighttime power. Double the battery bank to accommodate that will slow down the rate of charge (to about 6%) which would affect daytime use.

The second system with 1860 Watts of panel could support about 600 Amp hours @ 24 Volts.

The first system has an awkward panel to battery capacity ratio and would work better with about 130 Amp hours. Unfortunately that's not one of the options.

mike95490

Re: Before I make ANOTHER mistake

South Africa wrote: »

Before I commit on these two system:

For both systems:
- 2 pole fuses for all incoming strings, as well as between batteries and controller with the correct amp fuses per holder.
.........

Sure you want fuses (that you can only check at night) If you pop a fuse out of a live fuse holder, you have made an arc welder,
fuses are not circuit breakers with an arc extinguishing flame chute.

Cariboocoot

Re: Before I make ANOTHER mistake

mike95490 wrote: »

Sure you want fuses (that you can only check at night) If you pop a fuse out of a live fuse holder, you have made an arc welder,
fuses are not circuit breakers with an arc extinguishing flame chute.

Probably not going to be much of a problem, even above 15 Volts. People usually don't have to pull fuses, especially not under full power.

South Africa

Re: Before I make ANOTHER mistake

Cariboocoot wrote: »

... 400 Watts of panel running 'full battery' (all power in and out) will manage about 1000 Watt hours total with 5 hours equivalent good sun. Double the battery bank to accommodate that will slow down the rate of charge (to about 6%) which would affect daytime use.

Being lights preferred, this is no problem for system 1.

Cariboocoot wrote: »

The first system has an awkward panel to battery capacity ratio and would work better with about 130 Amp hours. Unfortunately that's not one of the options.

System 1 batteries is not cast in stone. Must be right, at the cost of new batteries. :-)
130ah - tell me more!? :-)

Cariboocoot wrote: »

The second system with 1860 Watts of panel could support about 600 Amp hours @ 24 Volts.

How fast will it charge on a good day? Ideally it would be nice to use lights at night, and be on full solar power by 10am, because batteries are charged. :-)
600ah @ 24 = 6 x 100ah batteries series / parallel?

South Africa

Re: Before I make ANOTHER mistake

Cariboocoot wrote: »

Probably not going to be much of a problem, even above 15 Volts. People usually don't have to pull fuses, especially not under full power.

It is going to be a problem. :-) Thanks mike95490!

It is daytime that I would probably want to 'switch off' the panels.
Same as between the inverter and batteries, if one wants to 'do' something.

That is fine ... I can change that.

Cariboocoot

Re: Before I make ANOTHER mistake

South Africa wrote: »

System 1 batteries is not cast in stone. Must be right, at the cost of new batteries. :-)
130ah - tell me more!? :-)

Well 400 Watts of panel ought to put out about 13 Amps on an MPPT controller (400 * 0.77 /24 = 12.8 ) on a 24 Volt system. So using the 10% rule-of-thumb that's 130 Amp hours. That balances with the 25% DOD rule-of-thumb for about 780 Watt hours DC. If you go deeper in DOD it takes longer to recharge.

How fast will it charge on a good day? Ideally it would be nice to use lights at night, and be on full solar power by 10am, because batteries are charged. :-)
600ah @ 24 = 6 x 100ah batteries series / parallel?

That is a very good question. Unfortunately charging isn't linear: no X Amps * Y hours = XY Amp hours. Every battery as a Peukart curve which affects both discharge and (recently discussed in another thread) recharge. The last 20% takes the longest, so to speak. Can you use it by 10:00? Depends on how early the sunlight hits the panels and soon they'll reach maximum power (direct sunlight).

One thing someone is bound to mention here so it may as well be me: with the 612 Amp hours made up of 102 Amp hour batteries you would have six parallel strings. That is less than ideal and may prove problematic in maintaining even current flow through all strings. It's not impossible, but it increases the possibility of trouble. You need bus bars to connect all positive and negative strings ends to, and equal length wiring on all strings as well. The idea is to reduce resistance differences between the strings as much as possible. You will also have to maintain them carefully, watching for imbalances (differences in cell SG) and increasing resistance (corrosion or loose connections). It is easier to keep strings balanced when there are fewer strings and/or more Voltage.

25% DOD on 612 Amp hours will be about 3 kW hours AC. In theory this could be 'harvested' in 3 hours with the 1860 Watt array. You're only looking at around 1kW hour overnight so that should be faster. There are a lot of variables that get in the way of making accurate predictions. Clouds, for example. Pterodactyls overhead. Low-flying UFO's. That sort of thing.

Cariboocoot

Re: Before I make ANOTHER mistake

South Africa wrote: »

It is going to be a problem. :-) Thanks mike95490!

It is daytime that I would probably want to 'switch off' the panels.
Same as between the inverter and batteries, if one wants to 'do' something.

That is fine ... I can change that.

DC disconnect, or cover the panels first.
Batteries, on the other hand, need switches.
Or a really brave heart.

South Africa

Re: Before I make ANOTHER mistake

For System 1, a battery monitor will be set to 25% DOD and then switch over the utilities for the (now renamed) lights only system.

And 130ah @ 24v is then what it will be. (2 x 12v 65ah batteries connected in series.)
And 780 Watt hours DC is about 740 Watt hours AC? So one needs to manage what lights are on at night, and why. :-).

Six parallel strings are too much. I like it to be KIS. Am I right in saying that a) 6 x 200ah batteries or b) go 48v?

When it all works lekker, then I can do that for more usage of stored power per day / night.

For the immediate future I think a total of 6 x 102ah is ok (306ah), is enough, and as you say, that last 20%, system can swap over to feed inverter whilst getting the last few UFO's into the 'kraal'. :-)

Cariboocoot

Re: Before I make ANOTHER mistake

Uh, two 65 Amp hour 12 Volt batteries in series is 65 Amp hours @ 24 Volts; no increase in Amp hours from series connection. You'd need four such batteries as two parallel strings of two in series.

Six parallel strings are too much. I like it to be KIS. Am I right in saying that a) get 4 x 150ah batteries or b) go 48v?

Yes. 48 Volts would automatically knock the strings down to three for the same capacity. Larger batteries would reduce the number of strings on 24 Volts.

Sometimes you have to balanced up working with what you've got against spending more money.

South Africa

Re: Before I make ANOTHER mistake

Cariboocoot wrote: »

Batteries, on the other hand, need switches.

I read on the site somewhere that some say yes for switches, others say no?
Rule here is FIRST switch off inverter, then disconnect the panels using the current breaker, then disconnect 1 pole using one eye to see what you are doing ... only then open the other eye ... :-)

Seriously, I will get a circuit breakers between controller and battery. Not too keen for a switch.

Cariboocoot wrote: »

Or a really brave heart.

My heart has taken one to many shock (AC and DC) to be brave anymore. :-)

Cariboocoot wrote: »

DC disconnect, or cover the panels first.

Panels are a wee bit high on the roof. Install once, leave well alone. :-)

DC disconnect = removing the cable, or flip the circuit breaker?

South Africa

Re: Before I make ANOTHER mistake

Cariboocoot wrote: »

Uh, two 65 Amp hour 12 Volt batteries in series is 65 Amp hours @ 24 Volts; no increase in Amp hours from series connection. You'd need four such batteries as two parallel strings of two in series.

Oeps, it is 2am here. :-)

Cariboocoot wrote: »

Larger batteries would reduce the number of strings on 24 Volts.

Sometimes you have to balanced up working with what you've got against spending more money.

Ditto, will save for 6 x 200ah batteries. With 25% DOD it is going to be a lot of power that I can use. :-)
Those Thomas Edison type batteries comes to mind. :-)

Cariboocoot

Re: Before I make ANOTHER mistake

South Africa wrote: »

I read on the site somewhere that some say yes for switches, others say no?
Rule here is FIRST switch off inverter, then disconnect the panels using the current breaker, then disconnect 1 pole using one eye to see what you are doing ... only then open the other eye ... :-)

Seriously, I will get a circuit breakers between controller and battery. Not too keen for a switch.

For occasional disconnects a circuit breaker can be considered a switch. If you can combine 'switch' with 'over-current protection' all the better.

My heart has taken one to many shock (AC and DC) to be brave anymore. :-)

Yeah that may be my problem too.

Panels are a wee bit high on the roof. Install once, leave well alone. :-)

Woos. My roof is 37 degree slope; so much fun to navigate!

DC disconnect = removing the cable, or flip the circuit breaker?

Best case: spring-loaded switch that snaps the power on/off (reduces risk of arcing). Otherwise ... panels can be covered to drop output to zero which makes removing wires safe. You can't do that with batteries though.

South Africa

Re: Before I make ANOTHER mistake

Cariboocoot wrote: »

Woos. My roof is 37 degree slope; so much fun to navigate!

BWAHAHAHAHA ... That I am ... see it is the height (having a granny flat on top), combined with my already starling track record that makes me wonder if climbing all the way up there is a good idea if I want to 'switch off' the panels. :-) :-)

And SWAMBO is keeping a hawks eye on me lately. Apparently she does not understand why I want to do it myself and not pay a professional. According to HER that will be cheaper. I have NO IDEA what she is on about. :-)

Cariboocoot wrote: »

Otherwise ... panels can be covered to drop output to zero which makes removing wires safe. You can't do that with batteries though.

WHAT!!! So throwing a blanket over the batteries does not stop them? But someone said throw a blanket over and it is safe ... !!! :-):D
(ROFLMAO - sorry, I just could not help!)

South Africa

Re: Before I make ANOTHER mistake

Man, I spend a few hours trying to work this out, going through all the posts. 2 steps forward 3 backwards. I just cannot get my head around this.

How would this work with the magic maths?
Daytime I power 1300w of equipment for as many hours as I can with a max set of 90% SOC, before switching power source.
At night 1190watt hours, then up to 75% SOC, assuming that nighttime starts at 90% SOC, as the system did not have enough time to charge.
First thing the next morning, batteries are re-charged before daytime load is added.

Assume 5.5 hours clear sunlight.
6 x 102ah batteries, 24v.
6 panels producing = 1860w array, 3 x series and 2 x parallel
Outback 80 controller
Phoenix 1600w inverter.

BB.

Re: Before I make ANOTHER mistake

One of the reason that charging a battery bank from 75% SOC vs 50% SOC...

If you have a 10% rate of charge--It would (in theory) take 2.5 hours to recharge a 75% SOC battery bank discharge. In reality, once you include Absorb stage, you need ~2 hours of 10% rate of charge plus ~2-4 hours of slowly declining Absorb charge--Or something like 4-6 hours of sun per day. That is about all a solar array can typically provide.

If you do a 50% discharge every day, then you need 5 hours of 100% sun, or maybe 3-4 hours of bulk charging and another 4-6 hours of absorb charging... Meaning you need something like 9-11 hours of charging per day--Maybe in full summer with a tracking array you could do that--But not for most of us.

From 0 to 80% SOC, flooded cell batteries will take near full current from the charging sources (10-13% rate of charge) without exceeding the absorb voltage... At ~80-90% SOC, the batteries will hit absorb set point and reduce their amount of current. Above 90% SOC, they will only accept a fraction of the charging current (5% to 2.5% Rate of charge or so).

You can play with the settings to try and increase charging (heavier than C/8 below 80% SOC) current. You can try higher Absorb voltage for part of the cycle (a few chargers will allow you to set, for example, a 15.0 volt bulk termination voltage, then fall back to 14.5 volts for balance of absorb), etc.

AGM batteries will take higher charging current than flooded cell (when bank is discharged). GELs (in general, available in the US) will take 5% maximum rate of charge--Which sort of kills their use in off grid solar power systems for daily deep cycling.

There are alternative charging cycles which can help... The cycle battery 50-80% daily, and recharge >90% every 7-10 days (one vendor says every ~28 days. If you cycle deeply Monday thru Friday (working) and take one or two Weekend days off (power usage wise), this type of cycling may be a viable option for your system (deep cycle during week, fully recharge every weekend).

Watch battery temperatures (hot batteries are batteries headed for a shorter life) and water usage (for flooded cell, refill ~every 2 months is probably ideal for many batteries).

Unfortunately, you may not know for 3-8 years if your charging/loading cycles are hurting your batteries or not. The joy of "cheap golf cart batteries" and banks that are not over-sized--Mistakes cost less.

-Bill

South Africa

Re: Before I make ANOTHER mistake

BB. wrote: »

If you have a 10% rate of charge--It would (in theory) take 2.5 hours to recharge a 75% SOC battery bank discharge. In reality, once you include Absorb stage, you need ~2 hours of 10% rate of charge plus ~2-4 hours of slowly declining Absorb charge--Or something like 4-6 hours of sun per day. That is about all a solar array can typically provide.

mmmmm ... the smaller the battery bank, the faster it will charge BUT Cariboocoot made me aware that I could damage the batteries having too many panels ... but I do not understand that yet, so now I will ask: Why? A MPPT controller will deal with that, won't it?

BB. wrote: »

From 0 to 80% SOC, flooded cell batteries will take near full current from the charging sources (10-13% rate of charge) without exceeding the absorb voltage... At ~80-90% SOC, the batteries will hit absorb set point and reduce their amount of current. Above 90% SOC, they will only accept a fraction of the charging current (5% to 2.5% Rate of charge or so).

Yes, system charges first the batteries, before moving over. Cariboocoot gave me an idea to set the battery monitor up to switch the inverter on again once it gets to 9X% charged.

BB. wrote: »

There are alternative charging cycles which can help... The cycle battery 50-80% daily, and recharge >90% every 7-10 days (one vendor says every ~28 days. If you cycle deeply Monday thru Friday (working) and take one or two Weekend days off (power usage wise), this type of cycling may be a viable option for your system (deep cycle during week, fully recharge every weekend).

Good idea!!!

BB. wrote: »

Unfortunately, you may not know for 3-8 years if your charging/loading cycles are hurting your batteries or not. The joy of "cheap golf cart batteries" and banks that are not over-sized--Mistakes cost less.

O such wise words !!!
Therein me being insistent on the least amount of 102ah Royal 'deep cycle' batteries ... till I get it right in the real world. :-)
At R890 = +-$85.23 per battery. :-)

-Bill

South Africa

Re: Before I make ANOTHER mistake

Maybe I must only have 2 x 102ah 24v battery on the system, to get to the sun power as fast as it can.

Light on a separate battery which are connected to a proper efficient charger that is powered by the inverter when it operates daytime?

I would think a charger would be more efficient than a solar panel and controller, extending the life of the batteries.

South Africa

Re: Before I make ANOTHER mistake

Any thoughts anyone?

BB.

Re: Before I make ANOTHER mistake

I am not quite sure I understand your question... From what I have seen, it is just as easy to "toast" a set of batteries with utility powered AC charger (continuous float charging, too high of float voltage, nobody checking electrolyte levels, not high enough voltage charging when deep cycled--such as a UPS application, people run UPS until dead with "small" AH AGM/GEL batteries which kills them in 1-2 years, etc.).

Vs solar, people run them until dead anyway (need electricity). We have read about renters/kids/guests that tear through a bank of batteries and kill them in a week (hair dryers, lights on and nobody around, long showers, TV+X Box video games and nobody on the couch, etc.).

From what I have seen, the only way for a long battery life is for the person that bought the battery bank to maintain the battery bank and monitor the charging system. And for that person to do a bit of reading on the various battery FAQs on the web and around here.

Certainly, if you have grid power and just need afternoon black out protection--It is difficult to justify solar panels. A smallish generator and some stored fuel can get you 5-10 days of emergency power pretty easily (if you can store the fuel or have backup sources like natural gas).

If your blackouts last weeks to months (after a major storm), then solar panels can be pretty nice. But they still are not cheap, and the batteries still need to be operated in their "comfort zone).

-Bill

South Africa

Re: Before I make ANOTHER mistake

Let me put it like this.

For 2 years now I have powerred, I realised today, a estimated 800-1000w per hour, dependant on clear days, via solar from 1330w panels using 2 x 102ah batteries to 'connect' the controller to the inverter. Average about +-23kwh per day in winter, no sun, summer we come close to -14kwh on sunny days. System start powering from about +-9h15am till about +-18h15 pm.

Recently I added lights. Now the batteries are not holding up so well, obviously, but more frustrating, also quite obvious, the charging in the morning takes to long. :-)

Splitting lights from the main system, needs a MPPT and a inverter.

Versus, if I can, go back to 2 x 102ah batteries but now on 1850w panels and a FM80.
Then I put the lights on separate 12v batteries, charging them using a http://www.ledsales.co.za/products/view_product/421/12V-20A-OMNIPOWER--HT-Series-Battery-Charger , using some of the 1850w power I generate each day via the 220v from the inverter.

I want to get away with as little batteries as possible, but I do not want to run into a problem I did not know about, around the next corner.

BB.

Re: Before I make ANOTHER mistake

South Africa wrote: »

For 2 years now I have powered, I realized today, a estimated 800-1000w per hour, dependent on clear days, via solar from 1330w panels using 2 x 102ah batteries to 'connect' the controller to the inverter. Average about +-23kwh per day in winter, no sun, summer we come close to -14kwh on sunny days. System start powering from about +-9h15am till about +-18h15 pm.

Just looking at the numbers... Assuming you get some pretty sunny days (6+ hours in non-winter/clear days) assuming well matched PWM or MPPT controller:

• 1,300 Watt array * 0.77 panel+controller derate = 1,001 Watt average "best case" noon time production.

Useful AC power per day (charge during day, discharge at night)--Assuming well matched solar array to battery bank (PWM or MPPT):

• 1,300 Watt * 0.52 end to end system eff * 6 hours per day = 4,056 = ~ 4.1 kWH per good day "useful" AC power

And for the battery bank, daily discharge should range from 25 to 50% maximum daily discharge (really suggest 25% maximum discharge for best operational battery life):

• 2 x 102 AH * 12 volts * 0.85 inverter eff * 0.25 max daily discharge = 520 Watt*Hours nominal daily discharge from battery bank
• 2 x 102 AH * 12 volts * 0.85 inverter eff * 0.50 max daily discharge = 1,040 Watt*Hours maximum daily discharge from battery bank

Recently I added lights. Now the batteries are not holding up so well, obviously, but more frustrating, also quite obvious, the charging in the morning takes to long. :-)

You have to match your system's output capabilities against the loads. Over discharging is very common. Charging issues (not enough hours of sun) are common too.

Splitting lights from the main system, needs a MPPT and a inverter.

Versus, if I can, go back to 2 x 102ah batteries but now on 1850w panels and a FM80.
Then I put the lights on separate 12v batteries, charging them using a http://www.ledsales.co.za/products/view_product/421/12V-20A-OMNIPOWER--HT-Series-Battery-Charger , using some of the 1850w power I generate each day via the 220v from the inverter.

My suggest is you are "over paneled" and under batteried (unless your "dark loads" are in the 500-1,000 WH per day range). The typical battery bank recommendation we would suggest for 1,300 watt array:

• 1,300 Watt *1/14.5 volt charging 0.77 panel+controller derating * 1/0.05 rate of charge = 1,381 AH @ 12 volt battery maximum bank
• 1,300 Watt *1/14.5 volt charging 0.77 panel+controller derating * 1/0.10 rate of charge = 690 AH @ 12 volt battery nominal bank
• 1,300 Watt *1/14.5 volt charging 0.77 panel+controller derating * 1/0.13 rate of charge = 531 AH @ 12 volt battery minimum "cost effective" bank

Your present system (if I got it right) is remendously over paneled for a pair of 102 AH @ 12 volt batteries.... A healthy nominal system for that bank would be on the order of:

• 2 * 102 AH * 14.5 volts charging * 1/0.77 panel+controller derating * 0.10 rate of charge = 384 Watt array "nominal"

Of course, if you use a lot of power during the daylight hours (such as water pumping/irrigation), you could make use of that larger array.

-Bill "my 2 cents worth" B.

South Africa

Re: Before I make ANOTHER mistake

YES!!!

Thank you for the calculations.

BB. wrote: »

... you use a lot of power during the daylight hours ... you could make use of that larger array.

That is EXACTLY what I designed.

Cariboocoot made a point somewhere that the array may damage the batteries, which made me worry. That is a unnecessary risk to take. But, a MPPT controller will not do that, will it?


Indeed we have a LOT of sunny days in Cape Town summer, with a winter rainfall to boot. There is an article in the RSSFeeds about SA's lovely sunshine. :-)

In December, Cape Town, sunrise @ 05:28, sets at 19:43 total of 14h 14m daylight, passing near directly over the panels, which are elevated for a clear airflow around them. :-)

System starts operating about 9ish in summer and keeps going till about +-18h00 when the SOC is 90%, switching back to grid.
So yes, a total of +-9 hours of no utilities power used.



Batteries are a rather huge part of a solar off grid system, and they require careful management, TLC and you are going to be replacing them in the future, no matter what you do to them.

Grid tie is the answer, if legal, you use it on the spot but you need to spend more funds to get anti islanding in place and then a way to supply the load if there is insufficient sun coupled with grid failure. So you would still need batteries. :-)

Now I can get cheap lead acid batteries for the few emergencies if and when, with some lights to boot. :-)

South Africa

Re: Before I make ANOTHER mistake

Do I have it right now?

Watt average "best case" noon time production: 1,850 Watt array x 0.77 = 1424.50 watt noon

Useful AC power per day, on a MPPT controller:
1,850 Watt array x 0.52 end to end system eff x 6 hours per day = 5,772 = ~5.7 kWH per good day "useful" AC power

Battery bank, daily discharge of max 10%:
2 x 102 AH x 24 volts x 0.85 inverter eff x 0.10 max daily discharge = 416 Watt x Hours maximum daily discharge from battery bank

And for best suggested battery life, suggested 25% maximum discharge for best operational battery life:
2 x 102 AH x 24 volts x 0.85 inverter eff x 0.10 max daily discharge = 1,040 Watt x Hours maximum daily discharge from battery bank

Typical battery bank recommendation for 1,850 watt array:
1,850 Watt x 1/29 volt charging x 0.77 panel+controller derating x 1/0.13 rate of charge = 377 AH @ 24 volt battery minimum "cost effective" bank

For a pair of 102 AH @ 24 volt batteries, healthy nominal system for that bank would be in the order of:
2 * 102 AH x 29 volts charging x 1/0.77 panel+controller derating x 0.10 rate of charge = 768 Watt array "nominal".

Cariboocoot

Re: Before I make ANOTHER mistake

Cariboocoot made a point somewhere that the array may damage the batteries, which made me worry. That is a unnecessary risk to take. But, a MPPT controller will not do that, will it?

A good MPPT controller can have its output current limited by programming, even though there's more power potential behind it. The down side is that limits the current for everything, including loads, because the controller has no way of knowing where the current is going. You can't set it up to limit current to batteries at 'X' but current to batteries and loads at '2X'. Although perhaps in future the Classic with the Whizbang Jr will be able to do this because it does know how much current is going to the battery. Program a battery current limit in and the controller could automatically reduce output to prevent exceeding that (as measured by the shunt) or raise to maximum if needed providing the battery current was not at its limit. I don't think it can do this yet.

BB.

Re: Before I make ANOTHER mistake

There can be several problems with "over paneled" battery banks.

One is that if you are charging a deeply cycled battery many hours during the day, over ~ C/8 rate of charge for flooded cell batteries can over heat the batteries, Hot batteries have lowered charging voltage, and the controller can see the depressed charging voltage as needing more charging--And you can get thermal run away. One help is to get a battery charge controller that a remote battery temperature sensor to directly monitor battery bank temperature.

Another issue that can happen with MPPT charge controllers. Some MPPT controllers will run "unregulated" every 5-15 minutes when "sweeping" the solar array to figure out the Vmp-array voltage (Pmp=Vmp*Imp). During this time, the controller can dump 100% of the array wattage into the battery bank. There have been times where this will take a battery bank over voltage (a 48 volt battery bank has been taken >72 volts on a couple systems).

There may also be an issue with some (few/many?) that the battery bank voltage is not stable with near 100% batteries and very high current charging sources... As the charger cycles (and perhaps during day time load cycling), the time constant of the controller's feed back loop may be too short to control momentary high current/high battery voltage situations (i.e., you cycle a water pump on a 100% charged battery bank, the load takes a bunch of DC current, the charge controller supplies maximum current from the array, the pump turns off, the controller takes a few seconds to cut back on the current output--During that time the battery bank over voltages and turns off/damages the attached DC loads/AC inverter/etc.).

How often this happens, and what brands... I don't know that I have a good handle on this. It is a pretty rare event (ant least we do not see it much here) and it just becomes one of those rules of thumbs (5-13% typical solar charging recommendations, around 20-25% maximum charging current; or C/5 maximum long term charging current and C/2.5 maximum charging surge current--and full batteries are less capable of absorbing high charging surge current--And full AGM batteries may be even less capable of "regulating" battery bus voltage with high charging current).

I do not believe that MPPT type charge controllers are less susceptible to surge/over current issues during charging--And may some may actually may be more susceptible to "uncontrolled" output high voltage issues.

When you "operate on the edge" of a system's capabilities, you have to accept that you may hit those limits and accept the consequences that others never see.

-Bill

South Africa

Re: Before I make ANOTHER mistake

I have never heard the batteries bubble or get hot the last year. Bar when the Phocos was incorrectly connected to them. :-)

Cariboocoot wrote: »

A good MPPT controller can have its output current limited by programming, even though there's more power potential behind it. The down side is that limits the current for everything, including loads, because the controller has no way of knowing where the current is going. You can't set it up to limit current to batteries at 'X' but current to batteries and loads at '2X'. Although perhaps in future the Classic with the Whizbang Jr will be able to do this because it does know how much current is going to the battery. Program a battery current limit in and the controller could automatically reduce output to prevent exceeding that (as measured by the shunt) or raise to maximum if needed providing the battery current was not at its limit. I don't think it can do this yet.

(facepalm)

But off course!!!

Now this is a problem ... but wait, I have a battery monitor and it needs a shunt.

Tell me if this makes 'sense'?

• When the batteries are charging in the morning, the battery monitor shows the 'I' in double (positive) digit Amps, with no load, as the batteries are being charged.
• After a few hours, it drops lower and lower till zero, when they are fully charged.
• Then the load is switch on.
• As expected, the current then shows double digits (negative) and a minute or so later, settles back to zero, give or take 0.5A either side during course of the day, because the controller is compensating for the current drawn, thinking it is the batteries that need charging.

Having a Zero current displayed I presume means the current in from controller is in perfect balance with the current drawn by the inverter?
And it is not via the batteries for the battery monitor registers no current being drawn from the battery via the shunt, as it would go negative on the display? (Controllers negative is connect between shunt and inverter.)

Or is this a case of, let say 20a in and out, that that is all through the system and is messing up the batteries?

Have ordered a controller temp meter.

BB.

Re: Before I make ANOTHER mistake

South Africa wrote: »

Do I have it right now?

Watt average "best case" noon time production: 1,850 Watt array x 0.77 = 1424.50 watt noon

Useful AC power per day, on a MPPT controller:
1,850 Watt array x 0.52 end to end system eff x 6 hours per day = 5,772 = ~5.7 kWH per good day "useful" AC power

Of course, for most people, 6 hours of sun per day is near average "best case" sun for summer... I typically use ~4 hours per day (at least in the US) as the system design point where it will supply that minimum amount of power for 9+ months a year without firing up the generator.

I always look at the monthly average sun to see how the loads fit the power usage. This link has way more US?world wide solar irradiation locations:

http://solarelectricityhandbook.com/solar-irradiance.html

For many, the maximum energy usage is winter (more lights, less sun). For others, it may be summer (water pumping for irrigation, fans, A/C, etc.). Energy usage is a set of highly personal choices.

Battery bank, daily discharge of max 10%:
2 x 102 AH x 24 volts x 0.85 inverter eff x 0.10 max daily discharge = 416 Watt x Hours maximum daily discharge from battery bank

We usually use 1-3 days of load support (no sun), and 50% maximum discharge. 2 days * 1/0.50 max discharge gives us 4x the daily load or 25% discharge as a nice "optimum" matching of battery capacity, battery surge capacity, charging capabilities, and battery life.

10% cycling daily cycling is a bit on light side (lead acid do like ~25% discharge at least once per month for good life). 50% per day discharge can be just fine for RV applications where you do not have a lot of room for a large battery bank, and weekend/seasonal use means that batteries will usually age out rather than cycle out (i.e,. 50 weekends of cycling or ~100 cycles vs 1,500 cycle life--Battery will die of "old age" in 3-5 years, not 15 years of weekend cycling).

And for best suggested battery life, suggested 25% maximum discharge for best operational battery life:
2 x 102 AH x 24 volts x 0.85 inverter eff x 0.10 max daily discharge = 1,040 Watt x Hours maximum daily discharge from battery bank

Typical battery bank recommendation for 1,850 watt array:
1,850 Watt x 1/29 volt charging x 0.77 panel+controller derating x 1/0.13 rate of charge = 377 AH @ 24 volt battery minimum "cost effective" bank

10-13% rate of charge is a "healthy" sized solar array to battery bank ratio. 5% is minimum for solar array--Suggested for lightly used used system, weekend cabin use, etc. Below 5% rate of charge, too easy to over discharge a battery (sometimes called "deficit charging", not enough charging current) and some battery Mfg. recommend 5% to 10% minimum charging current.

For a pair of 102 AH @ 24 volt batteries, healthy nominal system for that bank would be in the order of:
2 * 102 AH x 29 volts charging x 1/0.77 panel+controller derating x 0.10 rate of charge = 768 Watt array "nominal".

Note there may be a question about your "2 * 102 AH x 29 volts" numbers... This would assume you are using 4x 102 AH 12 volt batteries. Two batteries in series (for 24 volt battery bank) and two of those strings in parallel (204 AH @ 24 volt battery bank capacity).

Before you where talking about using only 2x 12 volt batteries... If so, then it would be "1 * 102 AH x 29 volts" in the equation.

Another minor issue... 29 volts charging--That is the peak Absorb voltage that I use for these charging equations for calculating a relatively conservative system (favoring a larger solar array to battery bank capacity ratio).

Marc/Cariboocoot tends to use a lower voltage (like 24 volts or 26 volts as "average" charging voltage on a deeply cycled system). Use the numbers you feel comfortable with--It will change the array size by 10-20% or or so (1-2.5% rate of charge adjustment). It is not a big difference and does not materially affect the overall system performance--Just a bit more or a bit less charging capacity. Any numbers that are within 10% are pretty much "the same" in off grid/solar power systems. Do not bother trying to get more exact--Too many variables/unknown details to waste a bunch of time to get more exact--And weather variations will eat that 10% difference any way (more or less sun from "20 year average sun charts").

I tend to "gloss" over these details. They serve to "confuse" new to solar power (can't see the forest for the trees)... Once they understand the math and overall system design cook book, then talking about the details usually make more sense when laid against the overall system (better able to see the complexities and not get lost in the minor details).

-Bill

Cariboocoot

Re: Before I make ANOTHER mistake

South Africa wrote: »

I presume that means the current in from controller is in perfect balance with the current drawn by the inverter, but it is not via the battery for the battery monitor registers no current being pulled from the battery via the shunt? (Controllers negative is connect between shunt and inverter.)

Or is this a case of, let say 20a in and out, that that is all through the system and is messing up the batteries?

Yep. Zero current to the battery means all load demands are being met by the PV via the controller.

But unless you have a way of limiting the controller's output you don't want to up the PV to handle greater loads because it will increase the potential current to the batteries during charging, possibly to unacceptably high levels.

If you increase battery and PV capacity alike it stays balanced. The only trouble there is reducing the 10% DOD of the batteries overnight to the point where you barely use them. This is a waste of battery of course, as they will fail from age just the same and you don't get your money's worth out of them.

Cariboocoot

Re: Before I make ANOTHER mistake

BB. wrote: »

Marc/Cariboocoot tends to use a lower voltage (like 24 volts or 26 volts as "average" charging voltage on a deeply cycled system). Use the numbers you feel comfortable with--It will change the array size by 10-20% or or so (1-2.5% rate of charge adjustment). It is not a big difference and does not materially affect the overall system performance--Just a bit more or a bit less charging capacity. Any numbers that are within 10% are pretty much "the same" in off grid/solar power systems. Do not bother trying to get more exact--Too many variables/unknown details to waste a bunch of time to get more exact--And weather variations will eat that 10% difference any way (more or less sun from "20 year average sun charts").

No, I use system nominal as a minimum Voltage benchmark. That way the batteries are not drained below 50% SOC (nominal Voltage under load guarantees this). It is also when the maximum current will be required: at the lowest SOC. By the time you get up to Absorb Voltage the batteries should not be demanding that 10% current rate. Loads need to be discounted unless they are known to be consistent and high; most applications have very low average loads that cycle on and off and can be easily compensated for by the 10% peak capacity.

If you are pulling that kind of current at Absorb Voltage then it is almost certainly loads and not battery charging. Calculating minimum array sized based on 10% Amps * full charge Voltage is likely to result in an array larger than necessary, and most people don't want to spend an extra dollar if they don't have to. As it is panels don't come in all sizes so you have to adjust the actual array Watts to fit what is available. Being able to tweak this up/down a bit is a bonus.

So 220 Amp hours @ 24 Volts:
22 Amps * 24 Volts (min) / 0.77 = 686 Watts. That will probably end up as 700 Watts when real panels are used. (Four 175 Watt panels.)
22 Amps * 28.8 Volts (Absorb) / 0.77 = 823 Watts. Right there a full 100 Watts greater than will actually be needed. (Six 140 Watt panels to make 840 Watts - potentially 27 Amps current. Or four 220 Watt panels making 880 Watts - etc.)

Reason for numbers in example cited: see my signature. It works.