Need help with 24 Solar Setup

[Dre123]

Hi,
While I'm searching for the ''perfect'' full size school bus to start the conversion, i would like to complete my sketching of the solar setup.

I have gathered around all the info that i could and asked for help on another forum but i wanted to ask what you guys think of it.


I live in Canada, temperatures range from -20C to 38C. I know
We expect 4-5 hours of solar irradiance on sunny days.

After conducting an energy audit, i have come up with 4947wh/day (that's the worst case scenario), realistically we are more around 4000Wh/day.


I would like this setup to be as cost and energy effective as possible.
I am aware that the battery bank should probably be bigger but i am limited by the budget.
The plan would be to double the batteries in the future if we see the need to.


I still have 2 things i am unsure of.
First, should i connect the negative bus bar to the chassis? Why?

Second, how realistic is the possibility of running a 940W mini split with the current setup?

Let me know what you guys think. Thanks

[Bert06840]

That's not bad, and not very different from mine.

1. You can put three panels in series per circuit, increasing efficiency. Your controller can take in 150V, you'll be under that even on the coldest brightest day.
2. I would not put that much current through an MC4. Even if you do two times three panels. I'd run two or three circuits down, and combine them on a busbar.
3. Make sure your breaker is DC rated. AC breakers don't break DC and definitely not at that current level.
4. I think your 24V-12V DC-DC converter is oversized. You may want to consider a 24V-13.8V DC-DC converter connected to a smallish lead-acid battery instead. That way you have huge surge current capacity, but not the expense of the large converter, and larger idle currents. You are very unlikely to draw kilowatts from the 12V bus, and if you do, you are doing something wrong.
5. You don't need the shunt meter. The Xiaoxiang BMS that you seem to be using has a good coulomb counter. I have a shunt and never even had to install it.
6. I don't know that specific inverter, but one of the big catches is inverter quiescent draw. It's hit and miss. Some take 15W, some 50 or 60, some of the integrated ones up to 85, and the low-frequency inverters 100+.
7. I don't think your battery bank is too small. I have 15kWh and could do with less, but I like it oversized obviously. Our usage was 2-3kWh per day without paying attention to anything, but no hot water or AC obviously from that power.
8. My experience with good mono panels and a Victron MPPT is that the difference between a sunny day and a gloomy day is about 5 to 1. On a sunny summer day, it produces about 6 hours net, on a gloomy summer day about 1.2. In your case, that would mean 10kWh per day on a good day and 2kWh on a rainy one. In the SUMMER. In the winter, you should expect days on end where you will not even break the 1kWh. Sure, you could pack a 30kWh battery, but it's more sensible to have a backup plan.
9. I did not ground my earlier system, because I had a backup to charge my 24V batteries from the 12V van at the time. I had to use it in the winter. The converter does not work if my batteries are grounded. In my new 24V bus, I will ground it, at the battery bank. It is safer, interference capacitors will zap you and then there is RFI. Ideally you also ground the HV side of the inverter but that is often not doable.
10. Mini split is probably not the biggest problem, assuming inrush currents are reasonable. But I don't think you need it. We never felt the need for AC in Vermont, Maine, and other areas of the Northeast. You won't run it for long off the battery bank, but since you only really need cooling when solar load is high, your panels will put out enough power to run it directly.

[Bert06840]

One more thing. If the mini split is important, change your 270W panels to 375W. They don’t cost that much more but you have over half kW extra when the Canadian Summer heat makes you retreat inside

[Bert06840]

And, looking at your schematic again - 270W class panels should be fused at 15A. The 20A is never gonna trip.

[Dre123]

1.
Yes i understand the concept of low current/high voltage being more efficient then high current/low voltage on short distances.

My reasoning behind why i chose to do 3 arrays of 2 panels in series was for the possible better capacity of collecting more energy in case of partially shaded panels and also that it might take less work for the mppt to transform 61.88V to 24V instead of 92.82V to 24V.
I also think it's safer to have only 2 panels connected in series in case of a panel failing and causing back feeding.
What do you think?


2.
Good catch, i forgot to check the limitations of the MC4 connectors.They look like they are limited to 20A.
By bringing the fuses inside the bus it will really also simplify troubleshooting!
I could use the same method used in the main Bus Bar? Copper bar with the ANL fuses.

But what type of connector should i use instead of MC4 to extend the length of the solar panels ?


3.
I've put the breaker there more to have the means of disconnecting the solar panels while working on the other components.


4.
They only offer 25A, 40A or 70A. Power consumption off load is 20ma/H.
40A could be ok to run all led bulbs, 2 water pumps, 2 maxxair fans, usb chargers and gps. Lets say 400w to run all that at the same time =33A.



5.
It was more for the fact that i don't want to rely on my phone, i like fixed displays for quickly check on the numbers. But i also plan on getting the optional screen for the mppt.
I just looked at the app review for the xiaoxiang bms, and its scary. They're is a lot of bad reviews. I don't know if i should look for a better bms...



6.
I also couldn't find this information

[Dre123]

Originally Posted by Bert06840

One more thing. If the mini split is important, change your 270W panels to 375W. They don’t cost that much more but you have over half kW extra when the Canadian Summer heat makes you retreat inside

The problem is if i have to run the AC during the hot humid nights when no sun is shining. I am afraid to wake up to a depleted battery.

[Rucker]

Nice level of detail, great research. That's what it takes.

I might have missed it but I didn't see lengths on your wires. AWG depends on amps AND length.

Other feedback/comments.

1. I couldn't find the exact MPPT device, there's a 150/70, so I'm wondering if you're looking at an old spec sheet for the controller.
2. Others here have commented it is not good to have a disconnect switch between the charge controller and the battery bank. Sounds weird, I know, but they have compelling arguments (related to power surge from the solar array).
3. Batteries-LiFePo? Consider adding heating elements because you live up north.
4. Is that a Victron SmartShunt? Hopefully yes? Consider the bluetooth version for ease of viewing. I see the other comment saying it's not needed because of your BMS...not really sure about that.
5. Solar panels. You may want to consider tilting frames because research shows folks up north lose as much as 25% of solar gain due to latitude. Others have made really good points in other posts, which I agree with.

[bubb, the real one]

Does your AC say it pulls 970watts AC?
If so then it will pull over 9000watts DC, inverting ac to dc multiplies the need by 10

[Bert06840]

Originally Posted by bubb, the real one

Does your AC say it pulls 970watts AC?
If so then it will pull over 9000watts DC, inverting ac to dc multiplies the need by 10

Sorry, that is completely untrue. Inversion (DC to AC) is usually around 90% efficient, not 10.

8000W of dissipation would melt your inverter into a puddle of metal in seconds.

[Bert06840]

Originally Posted by Rucker

Nice level of detail, great research. That's what it takes.

I might have missed it but I didn't see lengths on your wires. AWG depends on amps AND length.

Correct!

But if he uses 10 Gauge wire, he can make quite a long run at 8 or 9A max before it makes any meaningful difference.

1. I couldn't find the exact MPPT device, there's a 150/70, so I'm wondering if you're looking at an old spec sheet for the controller.

Victron indeed makes that one. I received my 150/70 just minutes ago, by the way!

2. Others here have commented it is not good to have a disconnect switch between the charge controller and the battery bank. Sounds weird, I know, but they have compelling arguments (related to power surge from the solar array).

Very good comment! You're right. Mostly because the charge controller is powered by the battery, not the PV.

3. Batteries-LiFePo? Consider adding heating elements because you live up north.

Yes, but the Xiaoxiang BMS have a low temp cutout sensor, most of them two. Your point is valid though, and the batteries should be installed INSIDE.

4. Is that a Victron SmartShunt? Hopefully yes? Consider the bluetooth version for ease of viewing. I see the other comment saying it's not needed because of your BMS...not really sure about that.

You don't. This thing was invented for SLA batteries, and in the time of BMSes with charge counters it has become redundant.

5. Solar panels. You may want to consider tilting frames because research shows folks up north lose as much as 25% of solar gain due to latitude.

True, but here I'd go the pragmatic way. Spend $200 on better panels and you can forget about the complex tilting affair that will take you a week to design and build!

[Bert06840]

Originally Posted by Dre123

The problem is if i have to run the AC during the hot humid nights when no sun is shining. I am afraid to wake up to a depleted battery.

Well, then you wake up to a depleted battery. No problem! Your BMS will shut it off when needed.

No harm will be done to your batteries IF YOU CONFIGURE IT CORRECTLY.

Did you say hot humid nights? Is there a Canada, FL?

[Bert06840]

Here’s my 150/70 that just showed up minutes ago.




Oh boy, it’s massive compared to the 75/15. On the internet they all look the same.

[Bert06840]

Originally Posted by Dre123

1.
Yes i understand the concept of low current/high voltage being more efficient then high current/low voltage on short distances.

My reasoning behind why i chose to do 3 arrays of 2 panels in series was for the possible better capacity of collecting more energy in case of partially shaded panels and also that it might take less work for the mppt to transform 61.88V to 24V instead of 92.82V to 24V.
I also think it's safer to have only 2 panels connected in series in case of a panel failing and causing back feeding.
What do you think?


2.
Good catch, i forgot to check the limitations of the MC4 connectors.They look like they are limited to 20A.
By bringing the fuses inside the bus it will really also simplify troubleshooting!
I could use the same method used in the main Bus Bar? Copper bar with the ANL fuses.

But what type of connector should i use instead of MC4 to extend the length of the solar panels ?


3.
I've put the breaker there more to have the means of disconnecting the solar panels while working on the other components.


4.
They only offer 25A, 40A or 70A. Power consumption off load is 20ma/H.
40A could be ok to run all led bulbs, 2 water pumps, 2 maxxair fans, usb chargers and gps. Lets say 400w to run all that at the same time =33A.



5.
It was more for the fact that i don't want to rely on my phone, i like fixed displays for quickly check on the numbers. But i also plan on getting the optional screen for the mppt.
I just looked at the app review for the xiaoxiang bms, and its scary. They're is a lot of bad reviews. I don't know if i should look for a better bms...



6.
I also couldn't find this information

I’d install the panels like so: ><, basically nested V’s.

Reliability isn’t a concern for panels, really. Unless you are running a real mission critical application where hot redundancy is mandatory, failures are a rare occurrence and you’ll notice when a panel is down.

Backfeeding is not an issue. Pretty much all panels sold today have a built in Schottky diode.

I think the conversion curve doesn’t bend that way (vin ideally close to vout). For a 150V controller, they need to use >150V MOSFETs. Higher currents, more losses. So 66% of the current and 150% of the voltage is almost certainly converting more efficiently than not.

[Rucker]

Originally Posted by Bert06840

Here’s my 150/70 that just showed up minutes ago.

Attachment 57680


Oh boy, it’s massive compared to the 75/15. On the internet they all look the same.

I believe you. That's what I said-I could find that one. The diagram showed a 150/60-maybe just a typo.

[Rucker]

Originally Posted by Bert06840

Correct!
[my comment about heater pads 'cuz he lives up north]
Yes, but the Xiaoxiang BMS have a low temp cutout sensor, most of them two. Your point is valid though, and the batteries should be installed INSIDE.

The heater pads are to keep the battery above a critical temp so the solar panels can recharge it on clear cold days. Yes, a low temp cutoff is needed; but without some way to keep the battery warm it will not recharge.

I like all the other points you made.

[rossvtaylor]

Originally Posted by Bert06840

Sorry, that is completely untrue. Inversion (DC to AC) is usually around 90% efficient, not 10.

8000W of dissipation would melt your inverter into a puddle of metal in seconds.

I think what Bubb meant was going from 12 to 120 volts...not necessarily from DC to AC. Watts are Watts...I believe the confusion there was regarding the current draw increasing by a factor of 10, when dropping the voltage from 120 to 12. You're absolutely right...but I see what he was trying to get across.

[joeblack5]

Not sure I understand your 280 amp fuse between the cc and your battery bank.. the cc is 60 amp?

In both our conversions we use small and cheap cc per string of solar. Redundancy on the road..

I agree with more parallel strings to avoid partially shadowing. In our next conversion we will run them all in parallel with an electrodacus bms and no cc.

Good luck.

Johan

[Bert06840]

Originally Posted by joeblack5

In both our conversions we use small and cheap cc per string of solar. Redundancy on the road..

That’s good advice, I decided to go against it because I wanted to have consolidated data. If I had time, I would do a distributed system (like the micro-inverters used in residential solar), and then aggregate the data in the digital domain.

Of course, you could go distributed, and if you still want to have an integrated capacity display, you could parallel them right at a current shunt. But then the shunt is the single point of failure.

So I’m going cold standby. I pull in all the panel feeds, and parallel them right ahead of the controller. If the controller conks out, I got another 15-amp Victron in a drawer that I can swap out.

I agree with more parallel strings to avoid partially shadowing.

I’m going to do an apples vs apples comparison. Since I have all the panel connections on the board, I can re-wire it from 3*2 to 2*3 in a minute.

[kazetsukai]

Originally Posted by bubb, the real one

Does your AC say it pulls 970watts AC?
If so then it will pull over 9000watts DC, inverting ac to dc multiplies the need by 10

Watts are a derived unit. You're thinking of amps. 12A 120V AC will be 60A 24VDC, roughly.



Thoughts to OP:

• Mini split: Even the 970W rating, it blows out your energy budget in a short few hours. Assuming you didn't add the split to that number?
• I would look to see if any low frequency inverters fit the bill for your electrical needs, they'll be much better for surge loads. PowerJack makes some nice 24V LF inverters for low cost, my "8000W" (realistically, 3500W continuous) 24V cost me right around $350 IIRC. They are also split phase- a plus if you want your panel to be wired like residential service, and intend to install 50A RV inlet. Just an option!
• Instead of using inline solar combiners, consider a PV combiner box, which will come with fuses/breakers and usually lightning arrestors/grounding equipment.
• Speaking of which, you show an AC panel and inverter but I don't see any RV style inlet. Intentional? It is quite convenient to have either a 30A (120V) or 50A (240V) inlet.

Looks pretty good overall.

[Bert06840]

Originally Posted by Rucker

The heater pads are to keep the battery above a critical temp so the solar panels can recharge it on clear cold days. Yes, a low temp cutoff is needed; but without some way to keep the battery warm it will not recharge.

I like all the other points you made.

In my case I decided to go against heated batteries, because it’s just an extremely unlikely case. For me.

It only makes sense if the bus is parked without human occupants for long periods of time, and *needs* to charge itself. The humans will be dead way before the batteries

If unsupervised winter parking were to take place, I’d have to winterize it anyway - the whole plumbing is susceptible to frost - I’d wheel the batteries into the house.

Even if - because reasons - I were unable to, the self discharge rates for my LFP batteries are so low that I could let them sit disconnected essentially indefinitely. Nothing would happen.

Even if they were allowed to cold soak, and I was unable to disconnect them, there’s still the low temp disconnect of the BMS.

Heater pads can be useful, but the complexity of doing heater controls from solar only (it makes less sense to do it from battery power), seemed not worth it.

I’ve lived 15 years near the Arctic circle, so my smug subconscious brain reacts to where most of Canada lives: “oh, that’s cute”. But I’m not going to say that out loud