diy solar

diy solar

Tell me about SolArk.

What product(s) did you use for string-level
If you mean string level as opposed to panel level then all you need is something that disconnects the panels. My Outback Skybox has an internal jumper that I connected to a big red button near my electrical panel to shut down the inverter. I am not sure how it shuts down the panels but the inverter goes offline so that the electrical circuits are not energized by the batteries when the first responders pull the meter or turn off the main panel. You may need to verify with your local jurisdiction that this is sufficient.
 

Attachments

  • RSD Button.jpg
    RSD Button.jpg
    83 KB · Views: 0
If you mean string level as opposed to panel level then all you need is something that disconnects the panels. My Outback Skybox has an internal jumper that I connected to a big red button near my electrical panel to shut down the inverter. I am not sure how it shuts down the panels but the inverter goes offline so that the electrical circuits are not energized by the batteries when the first responders pull the meter or turn off the main panel. You may need to verify with your local jurisdiction that this is sufficient.
Yep, that is inexpensive! Unfortunately ... 2014 NEC ...

690.12(1) through (5) A short requirement, only six sentences.

PV system circuits installed on or in buildings shall include a rapid shutdown function that controls specific conductors in accordance with 690.12(1) through (5) as follows.

  1. More than 5’ inside a building, or more than 10’ from a PV array
  2. Controlled conductors shall be limited to not more than 30V and 240 volt-amperes within 10 seconds of rapid shutdown initiation.
  3. Voltage and power shall be measured between any two conductors and between any conductor and ground.
  4. The rapid shutdown initiation methods shall be labeled in accordance with 690.56(B).
  5. Equipment that performs the rapid shutdown shall be listed and identified.
During the day, even if disconnected, any PV wires will be carrying voltage potential wherever they run ... if standard operating procedure for a fire department is to shut off power to a burning building, then it sure would make sense for them to be able to turn the PV wires off as well. So the 2014 version of this makes sense to me ... the module level stuff, not so much ...

Your Big Red Button will certainly be a part of my system, but I'll look for one that can turn off some remote switches ... thanks much.

By the way, I forgot to mention ... I'm going through my roof into my attic to a dedicated Inverter Battery room in my house ... you might be running everything outside your house ...
 
With NEC 2014, look to me like all you have to do is extend array of panels to edge of roof, and put a disconnect switch on the wall within 10' of the array. Inverter within 10' also does it. So could just put inverter on roof next to panels if on wall within 10' doesn't fit for some reason.

No per-panel RSD needed.
 
Here in So Cal, we do need to meet NEC 2017 now, including per panel rapid shut down and arc fault detection and shut down. For my small array, t nearly doubles the cost. Te more panels, the less the burden, because you typically only need a single arc fault an RSD device, with just a fairly cheap module at each panel for the shut down. Just learned that the master RSD and arc fault unit from Schneider runs about $685 and then I also need to buy the Tigo boxes for each panel.
 
I I was thinking you might be able to connect two lower voltage panels in series to a single RSD. But maybe not low enough voltage panels?
At one point I saw 4-panel RSD. I figured with my "12V" 120W panels having about 21 Voc, I could connect 3 in series, four string of 3s on one RSD. Then I'd only need two RSD for one 24 panel, 3000W string. Haven't seen that model in a while.

If you have two "24V" panels, likely they exceed 40 Voc each, so probably doesn't work for you.

One of these panels is 85 Voc


I guess max wattage under 80V is the thing to look for. Like 360W, 70V. But need a deal too.


But watch wattage of the RSD, too. Some can't take the higher wattage panels.
 
The 80 volt limit is for Arc Fault. If the DC open circuit voltage exceeds 80, then Arc Fault protection is required. So you could get away without arc fault if all the panels are in parallel.

But RSD has to drop the voltage of any wires more than 1 foot from the array to below 30 volts. There are not many panels that are that low.

Tigo has one RSD device that takes 2 panels. It isolates both in the off state to reduce the voltage to that of a single panel. Even a single panel can be well over the 30 volts, but that is acceptable as long as no hot wires extend past the solar array. Enphase has demonstrated that the voltage of a single panel is not very likely to make and sustain an arc. I think this is why ARC fault is systems are only required if DC is over 80 volts.

Going further in the rules... It does state "under 30 volts outside of the array", but since most single panels can be higher, the rule now also states "voltage must be under 80 volts within the array". That pretty much means each panel must be isolated.
 
This one says 2017, conductors within array boundary must be < 80V as part of RSD.
Maybe also further away 30V? Got a link? I thought that was 2014, done away with in 2017.

Maybe this allows panels < 80V within the array, isolated from wires outside the array.


Oh, here's a link with 30V also:


So wires outside the array either have to disconnect from MPPT capacitors, or have capacitors discharged.
 
With NEC 2014, look to me like all you have to do is extend array of panels to edge of roof, and put a disconnect switch on the wall within 10' of the array. Inverter within 10' also does it. So could just put inverter on roof next to panels if on wall within 10' doesn't fit for some reason.

No per-panel RSD needed.
Interesting ... here's my current layout ... I was just about to pull the trigger on buying 16 two panel Tigos until I read this (can't find any series level disconnects). The problem for me is that I can't put conduit on my roof so input C will be problematic (A,B,C,D are each 2p4s arrays, 164V, 22A). I had been planning on dropping the PV wire into the attic through Soladecks, then MC THHN to my inverter "room" (converted attic space SW of the chimney, access through the ceiling of my porch). Instead, I could drop the PV through my eves and run conduit around the house ... except for C ... for which I could stay with the original plan ...
1644099610105.png
 
Here in So Cal, we do need to meet NEC 2017 now, including per panel rapid shut down and arc fault detection and shut down. For my small array, t nearly doubles the cost. Te more panels, the less the burden, because you typically only need a single arc fault an RSD device, with just a fairly cheap module at each panel for the shut down. Just learned that the master RSD and arc fault unit from Schneider runs about $685 and then I also need to buy the Tigo boxes for each panel.
As Microinverter prices continue to drop, that architecture is going to be the better and easier way to meet RSD requirements…

Of course, the other aspect is tapering of AC-coupled power when off-grid, but the frequency-shift capability seems to now be pretty much ready for prime-time (and integrated into more and more hybrid inverters).
 
The AC coupling aspect is working perfectly.

My original system is 16 x 300 watt panels with Enphase iQ7 microinverters. They are on the essential loads panel on the output side of my Schneider XW-Pro inverter. I have only had 4 grid outages, and only one lasted a full day, but that was enough to see the system works perfect off grid. The only odd issue I ran into was the XW-Pro maximum battery charge current was dialed down to just 18% or just over 25 amps. I was not home when the grid went out, so the system did it's thing. The transfer relays opened, and the XW stated supplying the loads. The glitch from the switchover caused 2 of the microinverters to see bad power, so they waited the 5 minutes to come back online. It was cloudy when the grid first failed, so I can see the battery going between discharging a bit to charging a bit for the first hour of the outage, but then the clouds cleared a bit and the sun came out. The Enphase inverters were then making way more power than the essential loads needed, and the charge rate quickly went up above the set 25 amp limit. The XW-Pro responded by ramping up the frequency as expected, but evidently the iQ7's didn't reduce their output fast enough and/or far enough so the frequency ended up shifting far enough that 5 of the iQ7's saw it as grid frequency out of range, but the other 11 kept running. The loss of the 5 was enough that the battery charge current fell below the 25 amp limit, so it ramped the frequency back down. The iQ7 inverters came back on, etc. and this repeated about 5 times before the sun fell too low to produce too much charge current.

Had I been here, I would have certainly turned up the maximum charge current to allow it to capture the extra solar energy. The system stayed running on batteries until 1 am when the grid came back up.

I have 3 reasons I want to add DC solar charging to my system now.

1. Mainly to fully automate the battery charging without any external control needed. Even if I get the external control all working perfectly like 400bird, the DC charging is a failsafe that will work even if the network crashes etc.

2. Efficiency. Each time I cycle 10 kilowatt hours in and back out of the batteries with AC coupling, I am throwing away a full kilowatt hour to conversion losses. As efficient as the iQ7's are, the XW-Pro needs to then take the AC and convert it back to DC to charge the batteries. The DC charge controller is a much more efficient way of putting solar energy directly into the batteries. And it nearly eliminates my need to even bother with charge rate control of the XW, because while the sun is up, the XW will just sit in standby with the Enphase running the house loads, and the charge controller charging the batteries to use in the evening. Charge rate will just be everything I can get from the DC panels.

3. Dark Start. Hopefully this is never a problem, but it is certainly a big potential pain in the neck. With only AC coupling, you can get stuck. Let's say the battery bank runs down to where the XW-Pro has to stop inverting in the middle of the night. What happens when the sun comes up? NOTHING! With the battery based inverter shut down from low battery, there is no "grid" so the microinverters can't turn on either. My current solution is leaving my low battery shut off set to still have 50% in the batteries. Then in the morning, I can fire up my laptop on battery, and set the low battery limit lower and get the system to start. But it will take a few manual steps. With the DC solar charge controller, when the sun comes up, it will just start charging the battery bank. Once the batteries climb just 0.5 volts, the inverter will fire up, and then after the 5 minute delay, the microinverters will also start up. No human intervention needed at all.

I truly believe that the combination of AC and DC coupling in the same system is the best of both worlds.
 
3. Dark Start. ... With the DC solar charge controller, when the sun comes up, it will just start charging the battery bank. Once the batteries climb just 0.5 volts, the inverter will fire up, and then after the 5 minute delay, the microinverters will also start up. No human intervention needed at all.

Even if you don't have a relay for load-shed of 100% of load like I do, having some SoC based disconnect of a discretionary load could help. Might avoid going dark, also keep the excessive load off when inverter does come back on. If loads exceed AC coupled + DC coupled it would just go back to dark and DC only.

My plan is electric heat only when on grid, A/C and dryer heating element (one leg of split phase) only at 80% SoC.
Ideally, disable refrigeration at night if off grid.
 
Honestly, we don't have many or long power outages at this point. So far only once did they cut power due to wild fire threat here, but it could happen more in the future. When no one is home, the house draw here is under 1,000 watts. My battery bank will run that more than a full day with no sun. So when I had the one long power failure, I did manual load shedding. I just turned off the stuff we didn't need. And when it looked like we might not have power going into the second day, I fired up my old generator and had that run a 600 watt charger to the battery bank and my refrigerator. I put over 1,500 watt hours back into the battery, and when we wee heading to bed, the power came back on at 1 am. I kicked the XW into charge mode at 1,200 watts while we slept, incase the power went back off, we would have a bit more in reserve.

If power failures do start becoming more of a thing, I might add a pair of load shed relays. But since my furnace, stove, over, and dryer are all gas, I really don't have big loads to shed. Probably the furnace blower, it can pull about 900 watts, the hot surface gas igniter spikes to 1,500 watts, but for less than a minute. My fridge is my biggest steady load. It's an old pig and I need to replace it. The compressor never seems to stop. And then I have3 PC's, and the Dish Network DVR. My PC is an 8th gen i5, when I am just browsing the net, this thing only draws 90 watts. Wow, I just heard my fridge compressor stop, how unusual.
 
My fridge is my biggest steady load. It's an old pig and I need to replace it. The compressor never seems to stop. ... Wow, I just heard my fridge compressor stop, how unusual.


Save the fridge for backup, though. In case you have to wait for a replacement "motherboard" to fix a fancy new one.
"Reliability" has a quality all its own.

Check the temperature, see if it is reaching low enough. My upright freezer was running continuously, letting things get mushy and worse.
I finally recharged it and it is OK for now. None of the others seem to have lost refrigerant.

If yours is in trouble, consider picking up someone else's discarded old type as backup. Nobody ever gets rid of a car because it runs too well, but they do with appliances.
 
Back
Top