Indoor micro inverters

[big]

Replacing roof. Considering solar power system while I'm at it.

Generally like the micro inverter approach, although mounting them on/with the panels is not ideal. Harsh environment for the electronics and somewhat difficult to replace.

Yes, I know most of the micro inverters come with a 25 year warranty. I also know designing and verifying the electronics actually live this long is a daunting task. 25 year warranty possibly a figment of the marketing department.

So my question is, is there a reasonable way to mount the micro inverters indoors? Seems like the big issue is cabling.

 

[Bluedog225]

One advantage of micros is reducing risk of DC fire. Placed inside, not so much.

Plus potentially lots of roof penetrations. Or long cable runs.

If inside, why not go with a regular string inverter. Why the micro route?

Or a roof mounted string inverter.

 

[AntronX]

I was thinking the same recently. My concern is having to go up on the roof to replace failed inverters. Ideally I would like to have no electronics on the roof. My idea is to have a roof penetration for each 6 - 8 panels, make PV wire extenders to place micros in the attic for ease of access. If average extender wire length is 10' per panel using 10 AWG wire then additional power loss is only 0.6%

 

[zanydroid]

You are signing up for more work and code compliance problems. It most likely will violate rapid shutdown code and may require more aggressive wiring methods for DC depending on voltage (EG metallic conduit). Rapid shutdown has strict distance rules defining the array, and rules are more relaxed within the array.

My compromise is to use multiport micros but put them grouped easy to service locations on the roof. Using jumper cables to extend to the micro location.

I would not put a string inverter on the roof, that seems strictly worse than multiport micro (available in the 2kW range now). Since the microinverter ostensibly had a harsh environment design goal.

 

[GXMnow]

I have Enphase iQ-7 microinverters. Actually 16 of them, and they have been running for 4 years now. They have not degraded in any measurable way, even though I see then exceed 140F (60C) on hot summer days here is Southern California. The 25 year warranty did help me choose them, and the much safer and simple wiring was a big plus.

Moving the inverters inside the attic might reduce their operating temp a little, but my attic still hits 110F ambient temp in summer, and there is less air movement. And now instead of a single 2 conductor cable from each array (2 roof holes in my case, my system would need a 16 pairs of DC cables coming through from each panel. That is a lot of holes. Even grouping each 4 panels together would be 4 holes, but now they are carrying DC power with no protection either. And in my house, with a truss style roof, getting to the inverters in the attic might actually be more difficult than just lifting a panel off the racking on the roof. My upper roof is a little more difficult to get to, but the installers had no issues with the right ladder.

In most areas, you also need to have RSD (Rapid Shut Down) at each panel anyways. The microinverter or power optimizer meets that need. Running DC cables directly from the panels through the roof probably would not meet the code.

 

[pvgirl]

For panels mounted to buildings, unfortunately in the US you are going to be required to have some type of electronics module at or near the panel to provide the rapid shutdown function required by the electric code.

 

[zanydroid]

Why do you like microinverters? They’re only objectively useful in situations like really small roof planes not fitting enough panels for good string inverter minimum voltage, or extreme shading that’s somehow still going to yield some decent solar production on a few panels at a time.

And note I’m not an anti microinverter person. I’ve installed a ton on my own roof because of extreme conditions as above. But it’s not necessarily rational to even have solar in that case

 

[zanydroid]

For panels mounted to buildings, unfortunately in the US you are going to be required to have some type of electronics module at or near the panel to provide the rapid shutdown function required by the electric code.

It just needs to be within the definition of array boundary, right?

 

[AntronX]

For panels mounted to buildings, unfortunately in the US you are going to be required to have some type of electronics module at or near the panel to provide the rapid shutdown function required by the electric code.

Only if max PV voltage is > 40V DC. If I extend each panel then max array voltage is Voc < 40V. In theory I should be able to place RSD any distance away from the panel.

 

[Bluedog225]

I was thinking a string inverter on the roof could be enclosed to protect it from the elements. But still a harsh environment. Though I’m surprised there is not more discussion of this.

 

[pvgirl]

It just needs to be within the definition of array boundary, right?

Yes which is 1 ft in all directions of the array, so you could be inside the roof, but not really far enough away to get to more climate controlled space.

 

[zanydroid]

Only if max PV voltage is > 40V DC. If I extend each panel then max array voltage is Voc < 40V. In theory I should be able to place RSD any distance away from the panel.

It’s 30V which is outside the VOC of 54 cell panels.

So RIP

 

[zanydroid]

Yes which is 1 ft in all directions of the array, so you could be inside the roof, but not really far enough away to get to more climate controlled space.

Ok, that’s why I was advocating for (and have installed this way for myself once) grouping them in a servicing friendly way within the array that’s compliant with RSD and manufacturer instructions. They have to be shared by the panels, so at least one panel will have to be lifted. I guess maybe you can use a lighter shade structure that is within the 1ft rule

The downside is more wiring and connector costs and planning. And some voltage drop at 40V ish which sucks

 

[pvgirl]

Only if max PV voltage is > 40V DC. If I extend each panel then max array voltage is Voc < 40V. In theory I should be able to place RSD any distance away from the panel.

It’s 30V which is outside the VOC of 54 cell panels.

So RIP

Unless you just want to use "12V" panels in parallel only, then you can avoid the need for rapid shutdown, and arc fault protection as well

 

[AntronX]

It’s 30V which is outside the VOC of 54 cell panels.

So RIP

Not anymore:

When the 2017 NEC went into effect, those direct-current PV conductors that lay outside the array boundary, and those that extend more than three feet inside a building, were required to be controlled to not more than 30 volts within 30 seconds of rapid shutdown initiation; and the 240-volt-ampere requirement, from the 2014 NEC, went away

Starting January 1, 2019, however, those direct-current PV conductors that lay inside the array boundary, and not more than 3 feet inside a building, were required to be controlled to not more than 80 volts within 30 seconds of rapid shutdown initiation. See Section 690.12(B)(2)(2) of the 2017 NEC.

NEC Rapid Shutdown Requirements and UL 3741

iaeimagazine.org

 

[zanydroid]

Not anymore:

NEC Rapid Shutdown Requirements and UL 3741

iaeimagazine.org

I was looking at 2017 right before posting. Your excerpt covers the 80v within array rule (which incidentally I think allows you to 2s with many cell counts into MLPE). The scenario here is outside array in many cases. 1-3 ft vertical distance into building is not that flexible.

 

[AntronX]

3 ft vertical distance into building is not that flexible.

I can install the micros directly between the roof trusses or inside metal enclosure right there.

 

[zanydroid]

Unless you just want to use "12V" panels in parallel only, then you can avoid the need for rapid shutdown, and arc fault protection as well

Are those generally UL listed for roof mount/compatibility with engineering requirements? If not then might as well violate code in a more convenient and efficient way and use standard size modules

 

[zanydroid]

I can install the micros directly between the roof trusses or inside metal enclosure right there.

It’s ambiguous as to whether 1 or 3 foot applies. It requires careful parsing of or, and, and commas.

Because there is a second clause about outside array boundary in a parallel section to the within array boundary and 3 foot rule.

 

[zanydroid]

I have previously considered stacking separate RSD between inverter and module to address this.

However in my case it got even more complicated because I was using microinverters with paralleled together ports which means the microinverters could potentially bridge voltage back out to the modules, which means you need two RSD, one at panel and one at micro, to guarantee the voltage drops to safe limits in all cases. And that raises the possibility the second RSD would not be able to get the power from the panel that it needs to work.

(Microinverters with non paralleled ports should be fine)

But now you still have a MLPE up on the roof. It’s just 1/10th the circuit complexity and similar labor to manipulate

 

[pvgirl]

Not anymore:

NEC Rapid Shutdown Requirements and UL 3741

iaeimagazine.org

The language in the NEC 2017 is confusing, as of NEC 2020 and 2023 things are simplified:

(1) Outside the Array Boundary​

Controlled conductors located outside the boundary or more than 1 m (3 ft) from the point of entry inside a building shall be limited to not more than 30 volts within 30 seconds of rapid shutdown initiation. Voltage shall be measured between any two conductors and between any conductor and ground.

(2) Inside the Array Boundary​

The PV system shall comply with one of the following:

1. A PV hazard control system listed for the purpose shall be installed in accordance with the instructions included with the listing or field labeling. Where a hazard control system requires initiation to transition to a controlled state, the rapid shutdown initiation device required in 690.12(C) shall perform this initiation.
   Informational Note: A listed or field-labeled hazard PV control system is comprised of either an individual piece of equipment that fulfills the necessary functions or multiple pieces of equipment coordinated to perform the functions as described in the installation instructions to reduce the risk of electric shock hazard within a damaged PV array for fire fighters. See UL 3741, Photovoltaic Hazard Control.
2. Controlled conductors located inside the boundary shall be limited to not more than 80 volts within 30 seconds of rapid shutdown initiation. Voltage shall be measured between any two conductors and between any conductor and ground.
3. PV arrays shall have no exposed wiring methods or conductive parts and be installed more than 2.5 m (8 ft) from exposed grounded conductive parts or ground.

 

[AntronX]

(1) Outside the Array Boundary: Controlled conductors located outside the boundary or more than 1 m (3 ft) from the point of entry inside a building shall be limited to not more than 30 volts within 30 seconds of rapid shutdown initiation.

I guess I am good then. As long as I keep the micros < 3' from roof penetrations.

 

[AntronX]

Another option is to place all micros from the same PV column on top panel for easiest access. This way only top panel needs to be unbolted and tilted up while pivoting on bottom rail. Otherwise getting access to middle row inverter would be much more difficult.

 

[zanydroid]

Another option is to place all micros from the same PV column on the top panel for easiest access. This way only top panel needs to be unbolted and tilted up while pivoting on the bottom rail. Otherwise getting access to middle row inverter would be much more difficult.

Yup, this is exactly what I did. Though I think it's debatable whether using the top rail and pivoting up is easier than putting microinverter on the bottom rail and sliding the panel up higher on the roof, thus obviating the need to hold up the panel. I think I installed both ways. Haven't had to service yet.

I was able to save some costs / extra failure points from jumpers from using multi-port micros. Since those already need a lot of jumpers.

You would need to lay out the microinverters in a way that's compatible with the trunk cable system design. I think you can always squeeze them closer together, even with fixed length trunk systems like Enphase.

 

[AntronX]

thus obviating the need to hold up the panel.

I would install 2 bolts on panel frame to grab onto bottom rail and act as a hinge as well to keep panel from sliding down and falling during install. Then it can be propped up with a 2x4. Good point on mounting micros on bottom rail. Saves wire too.