Off grid design

[G005E]

Greetings, I'll try to be brief but I'm typically awful at that.
I've previously posted on the diy forum and naz forum but I figure it's time to go to the source! I've been lurking in here for about a month now.

We are off grid, building our homestead. We are still building our home so it's tough to do an energy audit. Based on historical data in a different climate, napkin math, and other comparisons I think I've isolated my usage to about 25kwh/day (I think I used the correct unit of measure) with potential peaks of upwards of 42k if a shop project is involved or other tools are used extensively.
I think I can continue to decrease this number, but will need to be living on site for adjustments to pan out. We are in AZ with good sun most year around and no plenty of unobstructed land for an array to have zero shade caused by vegetation or outbuildings

So far my plan was dual inverters (mainly redundancy but also in the event I need that much juice) 9-13k array and roughly 40kwh in Lifepo4 batteries and a 8.5k propane genie for backup/charging/bad weather.

I was all set on a pair of schneider 6848s until I learned the history of Bob and Robin amongst others and you guessed it, here I am.

I try and read about the Rosie everyday but being that it's still in infancy I can't gain much knowledge on my own. I was told there are people beta testing dual Rosie's. Is this going to be available for customers soon?

I also (likely my own ignorance) haven't been able to determine how much PV input the Barcelona, hawkes bay, and/or Rosie can handle. If I had dual Rosie's would I need dual charge controllers? Seems maybe with the hawkes bay and maybe not with the Barcelona as it has dual MPPT?

Is not having closed battery comms a big deal? It seems the eg4 batteries are wildly popular, but I also read a comment last night that I don't know the validity of, but it went along the lines of (6 eg4 batteries only output 20a overnight off grid as they are parallel). Is that true? That sounds awfully limiting on energy output overnight.
If that is the case, would the 220ah/304ah trophy batteries also suffer from such?

I apologize in advance if these are mundane questions or even repetitious. I've tried to search up and down the depths of the internet prior to asking and I just feel stuck.

[ClassicCrazy]

No not true on the batteries. You could take a lot more than that out of them. You can discharge them at 100a each but most people would never draw that much from just one. If you had 5  of them in series they would split the load and  then 20 amps each would equal 100a .
For every hour you draw 100 amp that would be 100ah .
The 48v eg4  rated at 100a  are about 5.2 kwh each in capacity.
 
There is a string calculator on Midnite site that you can put in some details of PV panels and it will calculate what the controllers can handle and how many you would need.
https://www.midnitesolar.com/sizingTool_barc-hb/index.php

Read Brians post - he is running a lot of power through his Hawkes Bay and has a Rosie hooked up too https://midniteftp.com/forum/index.php?topic=5441.0
Larry 

[Wizbandit]

Stacked Rosie's are a beautiful thing...
Operating off-grid you can set the threshold for the slave and it will sleep until needed.  Works perfect.  Grid Support mode will zero the incoming AC and power the loads until a set SOC or VBATT is reached.

[boB]

I was all set on a pair of schneider 6848s until I learned the history of Bob and Robin amongst others and you guessed it, here I am.

Awwww  Thank you !

That be us !  Welcome to the forum here.

I do frequent the NAZ forum too so might run across your post(s) there as well.

boB

[G005E]

No not true on the batteries. You could take a lot more than that out of them. You can discharge them at 100a each but most people would never draw that much from just one. If you had 5  of them in series they would split the load and  then 20 amps each would equal 100a .
For every hour you draw 100 amp that would be 100ah .
The 48v eg4  rated at 100a  are about 5.2 kwh each in capacity.
 
There is a string calculator on Midnite site that you can put in some details of PV panels and it will calculate what the controllers can handle and how many you would need.
https://www.midnitesolar.com/sizingTool_barc-hb/index.php

Read Brians post - he is running a lot of power through his Hawkes Bay and has a Rosie hooked up too https://midniteftp.com/forum/index.php?topic=5441.0
Larry 

Thanks for the feedback on the batteries. I thought the post I had come across was mistaken as it was the first I'd heard such, but wanted to be sure.

[qrper]

I was all set on a pair of schneider 6848s until I learned the history of Bob and Robin amongst others and you guessed it, here I am.

Yeah....Me too! Drank the Koolaid, and purchased the Rosie.

Mike

[ralph day]

Koolaid is tasty and nutritious!

[G005E]

Stacked Rosie's are a beautiful thing...
Operating off-grid you can set the threshold for the slave and it will sleep until needed.  Works perfect.  Grid Support mode will zero the incoming AC and power the loads until a set SOC or VBATT is reached.

That looks awesome. Can you shed some light on what the current configured set up is? Is that stacked Rosie's and a Barcelona with the breaker box?

[G005E]

Greetings again! So a follow up question I have is, what is the best design to run stacked Rosie's? Can I use one Barcelona for this or is that Ill advised?
If I was to run dual schneider it seems the most common way is to run with their box/cable management, Pdp, and then seems dual Mppt 100s. Alternatively it seems common to run a pair with a pair of classic 150s
I'm still trying to digest this new world, but it seems 2 Rosie's and 1 Barcelona if feasible would likely net me some potential cost savings without too many compromises

[ClassicCrazy]

As far as the Barcelona compared to Classics , you are better off with the Barcelona ( or Hawkes Bay) if you are going to have a large pv array since you can series them all up to around 500 to 600 volts ( use string calculator to determine cold temp compensation for voltage open circuit).
The Classics are limited to 150 to 250 volts depending on model.
The Rosie and Barcelona will share the same communications protocol which is canbus - the Classics are modbus . So easier for Barcelona and Rosie to communicate and both use MNGP2 .
Larry

[G005E]

As far as the Barcelona compared to Classics , you are better off with the Barcelona ( or Hawkes Bay) if you are going to have a large pv array since you can series them all up to around 500 to 600 volts ( use string calculator to determine cold temp compensation for voltage open circuit).
The Classics are limited to 150 to 250 volts depending on model.
The Rosie and Barcelona will share the same communications protocol which is canbus - the Classics are modbus . So easier for Barcelona and Rosie to communicate and both use MNGP2 .
Larry

Thanks for the insight Larry! I guess where my current confusion lies is: the Barcelona has dual MPPT. According to the string calculator I could run Canadian Solar 385w hiku panels in 8s2p to each channel and be well within VOC limits and end up with a roughly 12k array
Alternatively, I could also run the 400w hiku6 panels which are rated at a lower voltage and run up to 13 panels on a single string. Which would total close to 10.5k in panels and within the specs on the Barcelona string calculator
I'm not married to these panels, they just seem to be fairly dependable as a company and I can get either model for roughly 45 cents/watt new. Makes it easy to do calculations between the two.
Can the Barcelona handle 10-14k of PV? I haven't seen info either way.
I have my temp ranges set from -30 to 40 Celsius, which is -22 to 104 Fahrenheit. I'm in northern AZ and our record low and high are about 5 degrees less than that on either side (I set my temps broader than even historical records and it'll be ground mount so I didn't account for roof heat radiating)
Looking at how the system would be designed it seems as though I could use dual Rosie's, backplates, e panels, alongside ONE Barcelona, breaker, arc fault and ground fault, 2-300 surge protectors for battery side, and 1-600 solar surge protector for PV side, 2 MNGP2 (one for each inverter? The Barcelona includes one?), and required cabling to tie it all in. And then run the batteries as one bank and the panels as one array, and with 800ah of battery bank and 10-13k in panels I should theoretically be able to utilize the majority of dual rosie power output.

Is this all accurate/did I do it right? Hah

[Robin]

You do not need two MNGP2's. The one built into the Barcelona is all you need. IT will talk to all the MidNite equipment. IT will eventually talk to Classics also.
You mentioned that you want two Rosie's for redundancy. That being the case, you may want two Hawke's Bays instead of one Barcelona although it will cost a bit more. Is this site to be inspected? If not then you may opt out for the arc fault and ground fault module (s). They are a great safety feature demanded by the NEC, but I am skeptical of how many systems out there actually have arcing going on. Then again, if you are not a seasoned installer, it may just be cheap insurance against things going bad in the wiring.
You mentioned two SPD300's for the battery side? You will need one SPD300AC for the AC input and one for the Ac output. You will need one SPD600 for one Barcelona or two Hawe's Bays. Red wire to one input and black to the other regardless if this is a Barcelona or two Hawke's Bays. The Barcelona needs separate inputs. You cannot yet combine for one really big array.
Two Classics would also be very good option, but it does not yet communicate with the MNgP2, so the other two charge controllers make good sense. Either way will work just fine though.
Hope this helps.
By the way, we are out of Rosie back plates until August. If you can find a couple in stock somewhere, you may want to grab them.

[G005E]

That was tremendously helpful Robin, thank you!

[G005E]

Hello again, I have been the unfortunate party to many more rabbit holes of solar readings in the last week or two and now have new questions hahaha

If one is offgrid, has ONE Rosie and ONE Barcelona with a large bank, is the constant available AC power 7k? I'm feeling extremely ignorant again hence my probing question.
I would really really prefer to run a tier 1 inverter. I assumed I would need two for times that the family is not quite as judicious on load balancing large items.

In this scenario though, can I pull more than 7k constant via battery supplementation, PV input, and/or generator input? I guess the inverter can only invert so much so it seems any additional PV would be diverted to charging batteries while inverter was outputting maximum if the array was say 10k?
Which leads me to believe supplemental load power above 7k would only be possible with generator assist?

If someone needed a 15 minute draw of 12kwh how would this most effectively be done with a Rosie designed system?

Separately, in the diy section it would appear that the new diy inverters are rebadged SRNE. Is this correct? There was posts a year or so back about the future release of a 10k diy split phase inverter from midnite. The srne one is currently available to the public, is it still on path for the midnite version to be released?
If so, would this provide me a more ample overhead in potential demand needs?

I realize it is not near as robust as the Rosie, hence the price discrepancy, but if one must make compromises and can still purchase equipment from a reputable brand, it seems worthy of consideration.
As it sits, my other obvious inverter solutions are much cheaper than a Rosie, although potential issues seem much greater as well

[Wizbandit]

I have a dual stack set of Rosies in my shop for testing.  With a single you get 3500 watts per phase continuous. When I had a single I would set of an overload with a 2500W load on L1 (My RV with A/C running at 2300W) and opening the rollup door also on L1.  The opener was a modified Sears 1/2HP garage door opener. This stopped when I went to the dual stack which gives you 7Kw on both L1 & L2 phases.

I was the original tester of the SRNE 3548 3500W AIO (All In One) inverter with 140VDC MPPT input.  I tested it for about 3 years here at my homestead in Florida, through 95 plus degrees heat in the back of my garage, through a dozen tropical storms and hurricanes. I had my whole house's 15 amp circuits on it, lights, ceiling fans, 15A outlets, cable box.  It never missed a beat, never crashed or shutdown during outages.  All the heavy circuits were on a stacked pair of old Trace SW5548's which also working flawlessly.  I have 2 central A/C systems, a 2 ton upstairs and a 3 ton downstairs. The 2 ton was the only A/C on the battery and it was on the Trace setup.

The SRNE 3548 shutdown with a fan fail alarm about 2 years ago and I bypassed it & jumped the circuits over to the Trace system where it stayed until last year when I got a SRNE 10K 120/240 unit to test. I moved the circuits back and also included the 3 ton A/C just for a real test.

The 10K SRNE started and ran my 3 ton A/C along with all the 15A circuits that were once powered by the SRNE 3548.  It's been running for almost a year now without a glitch.  Midnite will probably not be re-branding the 10K units yet, at least not until a warranty department can be setup and we can deal with the failures in-house.

The biggest issue with the SRNE re-branded to MNS DIY is Midnite did not do enough forward thinking for warranty & support.  There was/is no DIY series repair bench setup, warranty was handled by shipping a new unit and piling up the bad units.  I had a lot of them shipping to my shop to inspect, very few suffered failures from manufacturing, most of them were abused by the DIY customer crowd not understanding how to properly size, wire and setup a solar power system.  They would just put a small battery and/or too small battery cables on the DIY3048 or DIY5048 and the strain would blow out the inverter's DC to DC section and throw a Fault-5 (DC to DC fail).  The DIY 3024's seem to suffer from a few cooling fan fail issues, not with the fans but with the internal fan sensor circuit which no one knows how to repair as we don't have schematics.  Plus, the labor to disassemble and reassemble is massive, it takes 45+ minutes to take a DIY3024 apart, if it's not burned then another 30+ minutes to locate bad parts and another 45 to 60 minutes to reassemble & test.  Really not worth repairing...