Solar power with grid as backup

South Africa · May 2014

Hi,

New here from Cape Town, South Africa.

This is what I have got, and how I put it all together. Idea is simple:
- If there is sun, then power the +-500w of computers, routers, switches and a fridge using all the panels can provide.
- At night, power all the house lights from same system, using the batteries.
- If there is no sun, then switch automatically to our grid / utilities company's power. Lights can only work off inverter.

I have got:
- 3 x Tenesol 310w panels @ 45.35v and 8.6a each connected to a TriStar 45 Controller.
- 2 x Tenesol 200w panels @ 26.7v @ 7.5a each, connected to a Phocos CML20 controller.
- Both controllers are connected to the battery bank to provide all the power that the panels can give when inverter operates.
- 4 x 102ah batteries connected as a 24v battery bank.
- Phoenix Victron C24/1600w inverter.
- Victron battery monitor installed in a Interlock to switch power between sources if batteries are drained of 10% of their power.

So, the panels and inverter powers a 3kva online UPS so that:
a) the inverter is protected from a overload when fridge starts up,
b) protect all the devices connected to it,
c) handle the 1 second switchover from inverter to utilities power.
Online is cool for they can handle a lot of drama from their input side, using their built-in batteries to ensure smooth output at all times.

To do the switching between utilities and solar, I had a switch built that uses a Victron battery monitor to measure that once 10% of the batteries are used, then system switches back to the utilities grid. Why? a) Utilities is cheaper than a huge battery bank, b) less batteries are required and c) batteries last longer.

Any suggestions on how to do things better?

Thank you in advance.

Cariboocoot · May 2014

Re: Solar power with grid as backup

Welcome to the forum.

I already know from another source that you are experiencing unreliable grid power there so we won't bother with the lecture on the cost of off-grid back-up power.

So let's look first at the batteries. Are these deep cycle batteries? If so you can take them lower than 10% before needing to switch power sources. If not they are probably not going to stand up to the task. I will presume they are 12 Volt, giving you 204 Amp hours @ 24 Volts. If that is the case you could discharge 25% and get about 1.2 kW hours from them.

You have a large amount of panel for that much battery. There's no advantage to this unless you are selling back to the grid when it's active. One array is 930 Watts and the other is 400 Watts. That would be 1330 Watts total and up to about 43 Amps peak current - approximately twice what the batteries need.

Another thing is the panel Voltages are odd. 45.35 is probably Voc, not Vmp. Vmp is important here. It should be about 35 to work with a 24 Volt system and it probably is 36 for those panels.

The other panels' Voltage at 26.7 is probably Vmp and is too low for a 24 Volt system unless the two are in series on an MPPT controller. Frankly you could leave these panels out. The 960 Watts provided by the others is more than enough to charge 204 Amp hours @ 24 Volts.

It sounds like you have some redundancy in the inverter department. Why do you have a UPS if you have an inverter? Normally we would use an inverter-charger for this sort of thing: critical loads placed on the inverter's output with grid supplied to its AC IN. Grid goes down, inverter takes over automatically. Not even a blip. When the grid is up it will provide power for the loads and recharge the batteries. Often with such a system no solar is needed at all. If your inverter requires one second to switch over get a better inverter.

South Africa · May 2014

Re: Solar power with grid as backup

Thank you Cariboocoot.

Yes, the power issues here SA are going to become a wee bit of a problem as the country is slowly growing up. Problem started 20 years ago when the then newly elected government decided to replace all the qualified experienced engineers, ignored the expansions plans back then and left the maintenance requirements aside. All the warnings where clearly communicated back then, but was ignored for over 10 years, that is until the power failures started in 2008. Will have to see what happens in the next 5 years and onwards. So far it is not too bad overall IF people and business head the warnings broadcasted on national TV. Having said all that, living in Africa is not for sissies and there are a lot of wind and solar farms coming online all over the country, with more planned.

Currently the batteries are Royals, the UPS 'deep cycle' batteries which I get for next to nothing. And with it not being a solar battery system, it is not a problem. Want to use the power as it comes in, not store it, bar for the lights at night.

Second consideration was in the mornings, the less battery usage the previous night, the faster they are charged and the sooner system switches over to 100% solar power.

All started with 2 x 200w panels a few years ago. Worked so well I then got the 310w panels because the 200w where not available anymore (today the 310w ones are not available either, stuck AGAIN!)

So I asked around and people said connecting them together via 2 separate controllers is ok IF battery charging is not the main issue. I used to have 2 x 35ah on the system, worked very well. Batteries then where purely for connecting the controller to the then 1000w Cotek inverter.

I must admit, I have not tallied up the load. I know the 1kva Cotek inverter at times had the red light coming on and back then I could not add the fridge. Today I can with the Victron. I also know that if I disconnect the 2 x 200w the system frequently switched back to grid power. 930w incoming is not just just not enough before 10am and after 3pm.

Attached the panels specs. I have been told to used the Open circuit voltage and Short circuit current as a safety net in all my calcs.

UPS is there to handle the 1 second break between inverter AC and grid AC, for system switches automatically between the two. If there is no sun AND no grid power, the batteries, which are not really being used, are then still fairly charged so I can continue for a number of hours.

Inverter/chargers are not the way to go for me. Grid AC is backup to the solar AC.

Cariboocoot · May 2014

Re: Solar power with grid as backup

Okay, a couple of things.

You use Voc and Isc for some calculations but not for all of them. In 'operating' mode the panels will produce Vmp, and if that isn't high enough to charge the batteries properly they don't charge. A Vmp of 26-ish certainly won't bring batteries up to 28-29 Volts which is generally what is needed to do the job right.

The low Amp hours to panel ration means that under some conditions such as extended outages the batteries may be subjected to more current than they can take. 43 Amps is over 20% charge rate on 204 Amp hours. That could ruin the batteries.

Your one second delay in switching AC sources shows that the inverter-charger would be the way to go. I think you haven't researched them properly. An Outback GVFX for example has a bidirectional AC IN and can even sell surplus power to the grid if that is allowed. More efficient than a standard OG system.

Solar does not produce AC. Inverters produce AC and they need DC to do it from. Normally you would have a battery bank sufficient to supply the power needs and solar to recharge it, with a back-up AC source such as generator or (unreliable) grid in case the solar was insufficient.

A refrigerator on its own can knock out a 1kW inverter due to the start-up surge. The Victron at 1.6kW is probably just enough. If there are more loads placed on it simultaneously it may not be sufficient.

If you get the batteries cheap enough you might as well increase the bank size to make use of the PV you've got. How much power it takes to recharge them is a matter of how much they are discharged, and a larger bank would simply be discharged less for the same amount of power.

A balanced system would normally use 25% of the battery bank per day and have a peak charge current of around 10% of the battery capacity. Such will operate quite well in 4 hours of good sun.

It sounds as though you are trying to use the batteries as a 'filter' on what would ordinarily be a PV direct GT system (only with an OG inverter). This is similar to what I do with my minimal bank and using as much power as possible during daylight hours in order to maximize the efficiency. In effect I need about 50 Amp hours to carry the overnight load. Oddly enough that is about 25% of the battery bank and the peak charge current is around 10%. It works fine if you get the numbers right.

I also use an inverter-charger and power AC IN from an inverter-generator if needed (i.e. no sun conditions). It works seamlessly; no delay in transfer at all and no need for a separate UPS system.

South Africa · May 2014

Re: Solar power with grid as backup

This is exactly what I am doing: Using the batteries as a 'filter' on what would ordinarily be a PV direct GT system (only with an OG inverter). Using as much power as possible during daylight hours in order to maximize the efficiency.

I use a OG inverter with solar charger and power AC IN from an utilities if needed (i.e. no sun conditions) and if there is no utilities and no sun, then, and only then, do I rely on batteries. It works seamlessly and if the batteries are drained, then we can watch the stars whilst waiting for the sun to rise. :-)

To simulate inverter/charger, which are relatively expensive here in SA, I had the attached built. I can adjust the 1 second to 10ms BUT the main use for the UPS is that when the system is happily powering the computers and related devices, when the fridge starts, the UPS absorbs the spike. Inverter is unaffected during those few seconds, and THAT is why I can use a cost effective 1600w inverter. :-)

FYI: Grid tie is illegal in most of SA and selling back to the utilities company is nowhere close to becoming reality but having obtained some info to date, the numbers they are talking is not going to make financial sense at all.

Few questions:
- Are the controllers not supposed to stop charging when the batteries are full?
- And whilst charging, do they not regulate the charge to not overcharge the batteries, irrespective of the panel wattage being so high?
- Can one mix 102ah and 105ah batteries?
- How many 102ah batteries can one add to the system using the 2 controllers?
- When the battery monitor says SOC is 100%, the current is at times zero being drawn from the batteries. Does this mean the system is well balanced?

I am contemplating changing a few things, to make the system more efficient.
1) Sell the 5 panels and 2 controllers
2) Get a FLEXMax 80a Outback controller.
3) With it maxing out on 2000w for a 24v system, inverter max of 1600w, plan is to get no more than 1650w of panels, and draw no more than 1300w from the inverter to not to push either device in order to make them last longer. My logic says if I use more than 1300w per hour, I am doing something wrong.
4) Add as much batteries as the system can charge comfortably.

How is my logic?

Cariboocoot · May 2014

Re: Solar power with grid as backup

South Africa wrote: »

Few questions:
- Are the controllers not supposed to stop charging when the batteries are full?

Yes. However charge controllers judge "full" by Voltage, maximum Absorb time, and/or End Amps. So far only (as far as I know) the MidNite controllers have a battery monitor addition that can end charging by SOC reading. After Absorb they will maintain a slightly elevated 'Float' Voltage to keep the load off the batteries if possible. Should load demands exceed PV capacity the batteries will make up the difference.

- And whilst charging, do they not regulate the charge to not overcharge the batteries, irrespective of the panel wattage being so high?

In terms of Voltage, yes. In terms of current only those that have programmable current limit will. And if the current limit is set at 'X' then that is what it is regardless of if the current is needed to supply loads or for charging the battery. Essentially it would limit the loads that could be run 'directly' from PV. So having batteries capable of handling the maximum PV output is desirable.

- Can one mix 102ah and 105ah batteries?

Yes. Although mixing dissimilar batteries is not ideal that much capacity difference would be little more than the variation you might find between two 'identical' batteries anyway.

- How many 102ah batteries can one add to the system using the 2 controllers?

With the estimated 43 Amps peak current you could have four parallel strings of 102 Amp hours. But beware of wiring them correctly as multiple parallel battery strings give rise to current variations between the strings which can cause trouble. See this wiring diagram link: http://www.smartgauge.co.uk/batt_con.html

- When the battery monitor says SOC is 100%, the current is at times zero being drawn from the batteries. Does this mean the system is well balanced?

That means the batteries are fully charged. In theory you can do this with very little PV (5% peak rate) or a whole lot of PV (>20% peak rate). Balancing is having enough to recharge in a reasonable amount of time while supplying average loads without spending a fortune on PV. I doubt any of us shut down while charging, so the PV will not only charge but also run most loads at the same time (except for surges and peaks). This is why the 25% DOD/10% charge ratio works so well in most cases.

I am contemplating changing a few things, to make the system more efficient.
1) Sell the 5 panels and 2 controllers
2) Get a FLEXMax 80a Outback controller.
3) With it maxing out on 2000w for a 24v system, inverter max of 1600w, plan is to get no more than 1650w of panels, and draw no more than 1300w from the inverter to not to push either device in order to make them last longer. My logic says if I use more than 1300w per hour, I am doing something wrong.
4) Add as much batteries as the system can charge comfortably.

Well if you consolidate to one array at 1650 Watts that would produce enough current for 530 Amp hours of battery. You could expect to draw about 1.2 kW 'directly' from the array once the batteries are charged. There is some variation on this because of changes in panel output due to temperature, wire loss, and conversion efficiencies. If you want to 'max out' an 80 Amp controller on a 24 Volt system it would handle about 2500 Watts.

Can you get something like a Kill-A-Watt meter there? http://www.solar-electric.com/kiacpomome.html You could measure your usage and get a better idea of averages and peak demands. Basically we are looking at "one outlet's worth" if this were in NA. You seem to use the European standard 230 VAC 50 Hz there and I believe there is a meter available for that.

I agree with you about the 1300 Watts; it's amazing how little electricity we can use if we try. Most of the time my system runs <500 Watts total, but I need the capacity for when the pumps have to run. One of them is 1200 Watts all by itself.

South Africa · May 2014

Re: Solar power with grid as backup

Yes, SA is 220v-240v @ 50hz.

What I did:
Step 1: Get an idea of your usage. Use the Efergy meter to see what is going on.
Step 2: Reduce the load i.e. LED lights, computers with lower PSU's etc.
Step 3: DO NOT cater for stoves, microwave, kettle, freezers etc. Not worth the system costs. Go gas, more efficient freezer etc.
Step 4: Now go solar with stuff that is on 24/7 and save 8-12 hours of utilities power. :-)

Cariboocoot wrote: »

With the estimated 43 Amps peak current you could have four parallel strings of 102 Amp hours. But beware of wiring them correctly as multiple parallel battery strings give rise to current variations between the strings which can cause trouble. See this wiring diagram link: http://www.smartgauge.co.uk/batt_con.html

This is my nemesis. I hate it when I connect batteries and the spark tries to electrocute me.

Is there a pic of "four parallel strings of 102 Amp hours"? I cannot seem to get my mind around it.

Cariboocoot wrote: »

That means the batteries are fully charged. In theory you can do this with very little PV (5% peak rate) or a whole lot of PV (>20% peak rate). Balancing is having enough to recharge in a reasonable amount of time while supplying average loads without spending a fortune on PV. I doubt any of us shut down while charging, so the PV will not only charge but also run most loads at the same time (except for surges and peaks). This is why the 25% DOD/10% charge ratio works so well in most cases.

Aha! With my magic box, the only thing that is on whilst morning charging takes place, is the inverter. No power is being drawn from it. Moment battery monitor says SOC is 100%, the inverter gets the load therefor I need the batteries charged as fast as is safely possible each morning so that the system can switch over from the grid to inverter power. At night all the LED lights are powered by the batteries.

Cariboocoot wrote: »

Can you get something like a Kill-A-Watt meter there? http://www.solar-electric.com/kiacpomome.html You could measure your usage and get a better idea of averages and peak demands. Basically we are looking at "one outlet's worth" if this were in NA. You seem to use the European standard 230 VAC 50 Hz there and I believe there is a meter available for that.

Have had one of these for the last 4-5 years: http://efergy.com/uk/e2-classic-2-0#.U2LRU-krhaQ

South Africa · May 2014

Re: Solar power with grid as backup

Four parallel strings of 102 Amp hours, is this how one would connect them?

Attachment not found.

Or would this be better?

Attachment not found.

For this is what I can do right now:

Attachment not found.

Cariboocoot · May 2014

Re: Solar power with grid as backup

Method #3 from Smart Gauge works best for four parallel strings: http://www.smartgauge.co.uk/batt_con.html

The first pdf of yours is like Method #4; it works, but is apt to cause confusion and therefor a mistake.

The second one is better, although the common connection points for the two pairs of strings should be one point, not two.

That image you posted looks suspiciously like one of my drawings. Did you find it on the forum or am I becoming anonymously famous?

South Africa · May 2014

Re: Solar power with grid as backup

Indeed, I found it on the forum. Thank you for your effort on it. :-) I took a sample, converted the values to metric. Now I understand better.

My mistake, yes, there is only one connection, like the 3rd pic. Thanks for the heads up re potential confusion. I have already destroyed 4 batteries being utterly completely stupid.

Right, I am then on the right track, thanks this forum.

Next question, to connect the batteries (note: I tend to swap them due to business needs, I need a quick disconnect and new connection), will this work?
I do not see more than 8 x 102ah for THIS system, starting off with 6.

Attachment not found.

Cariboocoot · May 2014

Re: Solar power with grid as backup

Anderson connectors are the default for heavy-duty current connections. Just be sure you get them sized for the maximum current expected; you don't want to use 50 Amp connectors where there is the possibility of 67 Amps flowing.

South Africa · May 2014

Re: Solar power with grid as backup

FLEXMax 80 controller can provide max 80amps if it is pushed with maximum panels on 48v.

Closest match is the SBS®50 Two Pole (110 amp). The Red one: http://www.andersonpower.com/products/multipole-sbs.html
Cable is therefor going to be 35mm @ 105amps. Amps there is 105amps max.
Cable lengths are max 1.5m (59 inches), with inner connecting cables 17cm (6,69 inches).

Believe it will be a good match with the controller, including going max on the controller one day.

Going back a few post.
Checked what the 2 x 200w panels are giving me. Turns out about 30v.
Whereas the 3 x 310w panels are giving over 40v in.

Do NOT ask me how we got there. All I see is that with the 2 x 200w panels connected, system starts to operate earlier and works later in the afternoon. Without them, I lose about 2-3 hours per day if I just use the 930w panels. :-)

Cariboocoot · May 2014

Re: Solar power with grid as backup

How exactly did you check these panel Voltages? They sound very much like Voc, which is not relevant to operation.

30 Vmp at the panels on a 24 Volt system is not good either, because by the time you get to the batteries you could still be down enough Voltage that they won't charge properly.

I don't know if you charge controllers are MPPT or PWM (the TriStar comes in either version). A PWM controller will only pass panel current; Voltage will settle at whatever Battery level is.

South Africa · May 2014

Re: Solar power with grid as backup

No MPPT. Already had a Phocos 100/30 die on me just 2 weeks after warranty expired.

Just measured the volts on the incoming wires, before controller (with the other controller disconnected), at about 10am this morning to see what is what.

NOW I remember!
2 x 200w are connect in parallel, to keep in the Phocos CML20's range.
3 x 310w are connect in parallel, to keep in the Tristar 45 range.

It worked to date, but it needs to be fixed. Therein the continious reference of the 1 x FLEXMax 80A with new MATCHING panels. Have to sell what I have. Anyone interested? :-)

Cariboocoot · May 2014

Re: Solar power with grid as backup

Shipping from South Africa might be a little pricey.

The 930 Watt array would be quite viable for a system. Nice fit to a MidNite Kid on a 24 Volt system.

South Africa · May 2014

Re: Solar power with grid as backup

Can scan it for you, then I email it and if you use a 3d printer, print them that side. :-)

Or move to Africa, we do have a wee bit more sun, especially in the Northern Cape. :-)

Seriously, so you recon the battery wiring, the connectors and the wires are now sorted? I am safe going for >100amps for cables and connectors?

If so, then I can start and fix that whilst I start shopping for a FLEXMax and new panels. Which is another thread for another day, and then the earthing of it all.

Cariboocoot · May 2014

Re: Solar power with grid as backup

I think wiring and connectors rated for 100 Amps will be sufficient even with a maxed out FM80 controller.

South Africa · May 2014

Re: Solar power with grid as backup

Baie baie dankie / thank you very much. I will not go forth and sort this out.

Next step is the disconnection of the panels.

Any feedback on this thread: http://forum.solar-electric.com/showthread.php?23346-Protecting-your-system-and-cables

South Africa · May 2014

Re: Solar power with grid as backup

Nice good rainy day today, so, as planned, little to no solar powered computers.

Batteries are charged comfortably from last nights usage though.

Because I limited the battery usage to 10% drain on sunny days, if the grid does fail today, which is unlikely, their is more than enough left over to power selected equipment for many hours if need be.

To have backup power for days like today is just not worth the cost of panels and batteries for we do not have enough days like this per year to justify it.

But, by reducing ones load, which makes it all cheaper, can make those sums viable.

South Africa · May 2014

Re: Solar power with grid as backup

Cariboocoot wrote: »

If you want to 'max out' an 80 Amp controller on a 24 Volt system it would handle about 2500 Watts.

This is what Outback says, for the FM80:
Attachment not found.

And using the +-80% rule (or as close as you can get to it) that I am advised by various suppliers in SA, I plan on doing this, if I can find the panels to add:
Attachment not found.

Higher voltage panels, that FM80 can handle with 24v system.

With initially 24v of 3 strings in series (6 x 102ah), later 4 strings in series (8 x 102ah) batteries on the system.

Safety, as discussed before, solar and battery breakers and battery cables and connectors all to handle min 150v @ 100amps.

Am I still on the right track?

Cariboocoot · May 2014

Re: Solar power with grid as backup

Well I know what Outback says about the FM80. If you look at normal operating conditions for 2000 Watts of panel you'll see it produces about 64 Amps peak. That's why you can actually 'over-panel' to 2500. Consider to that in hot temps such as you may experience in South Africa the array Voltage will be lower than normal.

I can't make out the second image at all. So I'll just mention that the array Vmp should not be more than 2X what would normally be used for the system: i.e. 70 Volts on a 24 Volt system. This usually negates any problems with high Voc in cold temps as well as keeping the array Voltage within a range that doesn't reduce the controller's efficiency significantly.

If you put in maximum array before adding the extra batteries you'll want to program a current limit: you don't want to whack 306 Amp hours of battery with 80 Amps of current (26% rate).

It's looking good so far.

South Africa · May 2014

Re: Solar power with grid as backup

Cariboocoot wrote: »

Well I know what Outback says about the FM80. If you look at normal operating conditions for 2000 Watts of panel you'll see it produces about 64 Amps peak. That's why you can actually 'over-panel' to 2500. Consider to that in hot temps such as you may experience in South Africa the array Voltage will be lower than normal.

Good point. But with warranties, I prefer to stay within the limits, having been down that road before, accidently off course. :-)

Cariboocoot wrote: »

I can't make out the second image at all. So I'll just mention that the array Vmp should not be more than 2X what would normally be used for the system: i.e. 70 Volts on a 24 Volt system. This usually negates any problems with high Voc in cold temps as well as keeping the array Voltage within a range that doesn't reduce the controller's efficiency significantly.

Attached it differently now.

Cariboocoot wrote: »

If you put in maximum array before adding the extra batteries you'll want to program a current limit: you don't want to whack 306 Amp hours of battery with 80 Amps of current (26% rate).

Then I best add all 8, if your sums are right. :-)

Using your figures (if I get it now): :-)
Max Array Watts: 2500w
Inverter % Efficiency: 0.77 (if efficiency is higher, more to 0.90, then it becomes a problem at 90amps.)
Average Watts: 1925 (2500 x 0.77)
System V: 24v
Amps Peak Current: 80.2amps (1925 / 24)
Battery Ah: 406 = (2 x 102 in series x 4 stings parallel = 24v @ 408ah battery bank?)
% of Panels: 23

Using my figures, do I have it right?
Max Array Watts: 1873 - as per attachment.
Inverter % Efficiency: 0.90
Average Watts: 1686 (1686 x 0.90)
System V: 24v
Amps Peak Current: 70amps (1686 / 24)
Battery Ah: 306 = (2 x 102 in series x 3 stings parallel = 24v @ 306ah battery bank?)
% of Panels: 23

Cariboocoot wrote: »

It's looking good so far.

At last! :-) All thanks to you. :-)

Cariboocoot · May 2014

Re: Solar power with grid as backup

Okay, how high up the mountain are you?

I run 82% efficiency @ 3200 feet.
Always remember the FM80 can have a current limit set below its maximum. You do not have to output 80 Amps even if the potential is there. That's one way to get better charging on less-than-perfect days.

Your array is somewhat confusing. It looks like you are running two panels in series per string. That makes sense. But you also say twenty-three panels? That doesn't work. With two in series there will always be an even number of panels. Nor will any number of 312 Watt panels (8.6 Imp * 36.3 Vmp) add up to 1873 Watts.

So two in series makes each string 624 Watts. Three strings is then 1872 Watts and six panels total. The array would be Voc 90.7 Vmp 72.6 Imp 25.8 Isc 27. Voc probably would never get as high as 119. That's 31% over.

You could add a fourth string for 2496 Watts. Current would go up by 1/3 and you'd have to plan the wiring for that.

Expect about 60 Amps peak current from the three strings. That's enough for all four battery strings (408 Amp hours) and then some. 14% charge rate in fact. You might still want to program an upper limit in.

South Africa · May 2014

Re: Solar power with grid as backup

Cariboocoot wrote: »

Okay, how high up the mountain are you? I run 82% efficiency @ 3200 feet.

Soon as I THINK I have it, another curve ball.

Way less than 900m from sea level.

Cariboocoot wrote: »

That's one way to get better charging on less-than-perfect days.

Now THIS is very interesting! I must try that one day. On hot days push the limit up and on colder cloudy weeks, lower it.

Cariboocoot wrote: »

Your array is somewhat confusing. It looks like you are running two panels in series per string. That makes sense.

Correct, max 6 panels as you say, or 6 parallel.
That means there are 3 sets of wires coming down, with circuit breakers on each, if I understood correctly.

Cariboocoot wrote: »

But you also say twenty-three panels? That doesn't work. With two in series there will always be an even number of panels. Nor will any number of 312 Watt panels (8.6 Imp * 36.3 Vmp) add up to 1873 Watts.

Only look at the green combinations, purple goes into the red, rest is just a standard template they use for all other types of panels and combinations.

Cariboocoot wrote: »

Voc probably would never get as high as 119. That's 31% over.

They make it safe for people like me that we do not have regular warranty disputes. :-)

Cariboocoot wrote: »

You could add a fourth string for 2496 Watts. Current would go up by 1/3 and you'd have to plan the wiring for that.

Warranty 1st, then experimenting later once I really got it all under control. Tired of getting my arse kicked. :-)

Cariboocoot wrote: »

Expect about 60 Amps peak current from the three strings. That's enough for all four battery strings (408 Amp hours) and then some. 14% charge rate in fact. You might still want to program an upper limit in.

Excellent. Well within my planned safety devices and cabling, yet with lots of spare, if I want to play. Without going to the manual, I presume I can set the limit from the front panel, and not another R5g's for that device they want to sell me to connect it to my computer. Nice idea, but silly price.

Wish they can make all of these controllers to connect to a Pc via a standard Cat5 network cable.

South Africa · May 2014

Re: Solar power with grid as backup

Cariboocoot, taking your advices from a earlier post of yours, are these numbers correct?

Good sun, average of 5.5 hours per day.
Batteries: 3 strings of 24v 102ah = 306ah

Both controllers connected to same batteries.
Note: Inverter is 'off' until batteries are fully charged.

Controller: Tristar 45
Max Array Watts 930
Inverter Efficiency 0.94%
Average Watts 874w (5.5 x 874 = 4808wh)
System V 24v
Amps Peak Current 36a
Battery Ah 306ah
% of Panels 12%

Controller: Phocos CML20
Max Array Watts 400w
Inverter Efficiency 0.94%
Average Watts 352w (5.5 x 352 = 1936wh)
System V 24v
Amps Peak Current 15a
Battery Ah 306ah
% of Panels 5%

I will get 12% + 5%, 36a + 15a for charging?

And for the UPS inverter for lights at night, will this be enough?
Usage of 200w x 4 hours = 800kwh per night
Max Array Watts 400w (2 x 200w Tenesol panels in series)
Inverter Efficiency 0.88%
Average Watts 352
System V 53v
Amps Peak Current 7a
Battery Ah 102ah (4 x 102ah in series)
% of Panels 7%

Cariboocoot · May 2014

Re: Solar power with grid as backup

This is what I get and how I get it:
930 Watt array @ 77% / 24 Volts = 30 Amps peak current
400 Watt array @ 77% / 24 Volts = 13 Amps peak current

43 Amps total current or 14% of a 306 Amp hour battery bank.

This assumes both controllers are MPPT type. If they are not then charge current is limited to the Imp of the panels.

Array has no effect on your nighttime power. If you don't store it or use it during the day it isn't there. Overnight you could pull 25% from that 306 Amp hour (76 Amp hours) bank and realize approximately 1.6 kW hours AC. Your array should be able to recharge that in a couple of hours providing concurrent loads aren't too much.

This brings up another interesting point: you may have 'surplus' power available from the array before the batteries are fully recharged because recharging isn't linear; as the batteries come up in Voltage their current demand will go down. It will serve you well to monitor that and calculate how much 'extra' power the panels could be making at any point in Bulk and Absorb. Then adjust loads accordingly. This may require some rethinking of your usage, such as leaving the inverter on or turning it on earlier and then adding loads to it as the power becomes available.

South Africa · May 2014

Re: Solar power with grid as backup

No MPPT's for these systems. But, I think I am going to advise my pal to get a MPPT controller. Will have to see in practice.
My MPPT comes into play once I get panels to match my 310w Tenesols after I bought the FLEXMax 80.

Cariboocoot wrote: »

Array has no effect on your nighttime power. If you don't store it or use it during the day it isn't there. Overnight you could pull 25% from that 306 Amp hour (76 Amp hours) bank and realize approximately 1.6 kW hours AC. Your array should be able to recharge that in a couple of hours providing concurrent loads aren't too much.

I will set the battery monitor to 25% use and see what happens.

Cariboocoot wrote: »

This brings up another interesting point: you may have 'surplus' power available from the array before the batteries are fully recharged because recharging isn't linear; as the batteries come up in Voltage their current demand will go down. It will serve you well to monitor that and calculate how much 'extra' power the panels could be making at any point in Bulk and Absorb. Then adjust loads accordingly. This may require some rethinking of your usage, such as leaving the inverter on or turning it on earlier and then adding loads to it as the power becomes available.

Excellent idea. Just that I have no idea how to make the Victron Battery monitor switch the relay to switch grid power off. At this point, that action happens when the SOC is 100%. Any suggestions?

South Africa · May 2014

Re: Solar power with grid as backup

For the 3rd system I am devising, as a introduction to solar, are these calcs now more or less in the ballpark, to recharge with 6% the batteries in 5.5 hours peak sun?
Assuming a non MPPT controller?
Usage of 200w x 4 hours = 800kwh per required night.
Max Array Watts 400w (2 x 200w Tenesol panels in series)
Inverter Efficiency 0.77%
Average Watts 308 (5.5 x 308 = 1694wh)
System V 53v
Amps Peak Current 6a
Battery Ah 102ah (4 x 102ah in series)
% of Panels 6%

And based on the above, for the usage required on a 48v system, which cost effective MPPT controller would one look at. I am ware of Tristar, Phocos, Morningstar available in SA.

See, if I can get the right combination here, I can sell him my 2 x 200w panels, IF they will do the job properly.

Cariboocoot · May 2014

Re: Solar power with grid as backup

South Africa wrote: »

Excellent idea. Just that I have no idea how to make the Victron Battery monitor switch the relay to switch grid power off. At this point, that action happens when the SOC is 100%. Any suggestions?

Without doing it manually ... it becomes complex. You'd need a computer to read SOC with some programed knowledge of the expected power curve to switch multiple relays for circuits to change them from grid to inverter. And of course back again if the load proved too much or the SOC dropped (as it would overnight).

My looking at the MX60 and saying "ah, time to turn on the water pump" is much simpler.

Cariboocoot · May 2014

Re: Solar power with grid as backup

South Africa wrote: »

For the 3rd system I am devising, as a introduction to solar, are these calcs now more or less in the ballpark, to recharge with 6% the batteries in 5.5 hours peak sun?
Assuming a non MPPT controller?
Usage of 200w x 4 hours = 800kwh per required night.
Max Array Watts 400w (2 x 200w Tenesol panels in series)
Inverter Efficiency 0.77%
Average Watts 308 (5.5 x 308 = 1694wh)
System V 53v
Amps Peak Current 6a
Battery Ah 102ah (4 x 102ah in series)
% of Panels 6%

And based on the above, for the usage required on a 48v system, which cost effective MPPT controller would one look at. I am ware of Tristar, Phocos, Morningstar available in SA.

See, if I can get the right combination here, I can sell him my 2 x 200w panels, IF they will do the job properly.

What you can expect from 400 Watts of array with all power being stored in and retrieved from batteries and 5.5 hours equivalent good sun: 400 * 5.5 * 0.52 = 1144 Watt hours AC. That should meet the 800 Watt hours (assume you did not mean 800 kW hours!).

Inverter efficiency isn't 77%. If it is, get a different inverter. The 77% derating is average panel output plus MPPT charge controller loss, so 400 Watts rating of panel in comes out as 308 Watts. Inverters are usually around 90% efficient. Some less, some more.

With a PWM controller you can only count of the current so you need to know the panel specs. A 200 Watt panel with a Vmp of 35 produces 5.7 Amps but a 200 Watt panel with a Vmp of 30 produces 6.6 Amps. And therein lies a caution about Vmp: it has to be high enough to charge the batteries. This goes along these lines: 12 Volt system Vmp 17-18, 24 Volt system Vmp 35-36, 48 Volt system Vmp 70-72. (Although a 48 Volt system may have a slightly lower Vmp as the wiring losses are inevitably less due to the high Voltage.)

You say four 102 Amp hour batteries in series. 12 Volt? That would be a 48 Volt system. Hence the caution about array Vmp. It would also supply roughly 1kW hour AC at 25% DOD so that is also good. The only possible issue is the panel Vmp and making sure the charge controller can handle 48 Volts (many PWM types can not).

South Africa · May 2014

Re: Solar power with grid as backup

Cariboocoot wrote: »

Without doing it manually ... it becomes complex. You'd need a computer to read SOC with some programed knowledge of the expected power curve to switch multiple relays for circuits to change them from grid to inverter. And of course back again if the load proved too much or the SOC dropped (as it would overnight).

I think this may just do it, manual says:

DF: Discharge floor. When the state-of-charge percentage has fallen below this value, the alarm relay will be activated. The time-to-go calculation is also linked to this value. It is recommended to keep this value at around 50.0 % for lead-acid batteries.

I set mine to 10%, and I just remembered why: Lights at night, 10 hours of lights.

Any case, the manual goes on to say:

ClS: Clear SOC relay. When the state-of-charge percentage has risen above this value, the alarm relay will be de-activated. This value needs to be greater than DF. When the value is equal to DF the state-of-charge percentage will not activate the alarm relay.

Rh: Relay high voltage. When the battery voltage rises above this value for more than 10 seconds the relay will be activated.

Rhc: Clear relay high voltage. When the battery voltage falls below this value, the relay will be de-activated. This value needs to be less than or equal to Rh.

So I recon when it gets to +-95/98% SOC, then switch the relay off, in other words, the grid connection is severed and the inverter powers things. Probably will need to play a bit to find the right balance.

Cariboocoot wrote: »

My looking at the MX60 and saying "ah, time to turn on the water pump" is much simpler.

I will be on the MX80 in the not too distance future, then I can join you. :-)

As a matter of fact, I want to see if I can add the 1.1kw pool pump to the equitation on the FM80, to start based on a) full sunlight and b) the SOC is 100%. Help of a tame electrical engineer comes to mind on this one. :-)

Solar power with grid as backup

Comments

Categories