Rail free mounting systems for standing seam metal roofs:Sunmodo and EcoFasten

[analogue]

I'm a newbie planning a 20kW DIY install solar array and happened to run across rail-free mounting systems for standing seam metal roofs. I have never heard of these - just happened to stumble across them while researching mounting hardware for the systems that use rails (Unirac and Ironridge).

Anyway, these seem relatively new and I wanted to get opinions from the experts on the pros and cons of using a rail-free system.

First a few details about my system and conditions:
- 20kW system
- Enphase IQ7+ or APSystems QS1 microinverters
- Location: Central Texas
- Roof pitch 2/12
- Temps: 100F+ for weeks at a time

The two rail free systems I came across
- Sunmodo SunDock - PDF Doc
- EcoFasten SimpleBlock PV - PDF doc

I haven't picked panels yet so I have flexibility to choose ones that are compatible.

Since I'll be doing all the labor, I'm most interested in the labor savings. The less the better of course

Other than that, price does also factor into things (systems with rails obviously have more parts and are more expensive).

The literature is short on cabling details and microinverter placement, maybe for a reason.

Any opinions on these or other rail free systems would be appreciated!

[solarix]

I say you should have a minimum of ~15 degrees (3/12 pitch) of tilt on solar panels so they will self clean in the rain.
Lower than this will cause PV panels to get dirty and stay dirty.

Tilting panels up will require a rail system and arraigning rows of panels with space between so shading does not occur.

The problem with standing seam clamps is they rely on the roofing fasteners to hold the uplift forces on the PV array. This is usually about double the fasteners than typical standing seam roof use. My local AHJ requires documentation of the fastener spacing so retrofitting with standing seam clamps is out.

I find rail-less mounting systems to be very frustrating to install in all but the simplest roofs. Figuring out where exactly they have to go in order to layout an array among vents and all the other constraints on a roof will put you back to the simplicity, versatility, and reliability of rails.

[bob-n]

I've only done one standing-seam install, so don't take this as an authority. In that case, it was a steep roof and having rails was a HUGE advantage in terms of safety. The lowest rail could be installed from ladders. Once that rail was installed, you had something secure to hold your weight while you put other things up on the roof, like panels. We also had safety lines and harnesses, but that one rail made the install so much safer. We used Ironridge with L-feet on S5 clamps.

[analogue]

Little more info on the site. The target area has no obstructions/vents so shading is not a problem at all. Also, the roofline is pretty high (~40ft at some points due do the sloping lot) and gets plenty of wind. The closest tree is at least 10+ feet below the roof line from what I can tell.

For those reasons, I'm not too concerned about dust build up given the pitch is 2/12 but maybe I'm underestimating things.

The picture is oriented north/south so can get an idea of the angle towards the southeast.

roof.jpg

[bob-n]

That's a great location. You'll make lots of power there.

2/12 is a relatively gentle slope. That's good.

On the unlikely chance that you need to get up there and repair something in the future, it will be helpful to give yourself at least 16" of walkway on the left, right, and top of the array. You can cover panels with plywood and lie on them that way, but you don't want to shimmy 20 feet across many panels covered with plywood to get to one bad connector.

As a rule of thumb, trees grow 1.5 foot per year. Some slower, some faster, but that's the general sense. Many of us plan to keep our houses and expect our equipment to continue going 20 years into the future, so if there are any trees that might get in the way later, it'll be easier and cheaper to take them down now.

[analogue]

bob-n, thanks for the tip on using plywood. Hadn't thought of that before. Also the recommendation for 16" clearance around all sides of the array was something I had not factored in at this point and I'm glad you brought it up

[J.P.M.]

Little more info on the site. The target area has no obstructions/vents so shading is not a problem at all. Also, the roofline is pretty high (~40ft at some points due do the sloping lot) and gets plenty of wind. The closest tree is at least 10+ feet below the roof line from what I can tell.

For those reasons, I'm not too concerned about dust build up given the pitch is 2/12 but maybe I'm underestimating things.

The picture is oriented north/south so can get an idea of the angle towards the southeast.

roof.jpg

Dust buildup can be a problem, particularly with low slope arrays, but in ways additional to those I'd thought of before installing my system. Most dust blows off or tends to accumulate in specific areas of panels and/or in a macro sense on sections of arrays, lower slopes --->>> more dust buildup. I've also found a bigger source of semi permanent fouling to be the dew that condenses on an array and turns the dust into caked mud. It doesn't blow away. It tends to build up. Low slope panels tend to turn into sandboxes, and the greater accumulation of dust, even though small, makes the caking worse for lower slopes as the mud runs down the array less. I note the bathtub ring on the lower 6"-12 " of many arrays and also note the generally lower width of the layer for higher sloped panels and vice versa.

FWIW, from what I think I might know, I'd be concerned or at least make some provisions for easier access to the array to be able to hose it off from the top 1X in awhile.

Also, w/no shading why are you using optimizers or microinverters ? More $$ or at least more to go wrong than w/a simple string inverter. The KISS principle applies. The higher probability of required service as system complexity increases is also another reason to consider easy access.

IF it was my project, I'd also run my plans past the AHJ before I bought anything and to check with them for any other requirements they may have (if any) such as for external loadings such as those incurred from wind or seismic that I might not be aware of. I sure don't know about your local codes, but I'd also check for required setbacks for first responder access. That array looks a bit tight. Also, if you use optimizers or micros, do you intend access to the panels in the middle of the array when service is required ? If so, how ?

Lastly, maybe I'm wrong, but it looks like the rail free systems you post don't allow much free space for ventilation under the array. That'll mean increased cell temps. and so lower system efficiency and so lower production.

Welcome to the neighborhood.

Take what you want of the above. Scrap the rest.

[J.P.M.]

That's a great location. You'll make lots of power there.

2/12 is a relatively gentle slope. That's good.

On the unlikely chance that you need to get up there and repair something in the future, it will be helpful to give yourself at least 16" of walkway on the left, right, and top of the array. You can cover panels with plywood and lie on them that way, but you don't want to shimmy 20 feet across many panels covered with plywood to get to one bad connector.

As a rule of thumb, trees grow 1.5 foot per year. Some slower, some faster, but that's the general sense. Many of us plan to keep our houses and expect our equipment to continue going 20 years into the future, so if there are any trees that might get in the way later, it'll be easier and cheaper to take them down now.

Might be a great location but not such a good slope. Besides being a good dirt catcher, the low slope of 2/12 will produce less annual output per installed STC W than a higher slope. I'd respectfully suggest neither one of those is inherently good.

Plywood or not, do not put weight on a panel by crawling on/over it. Besides even a remote probability of electrical shock/worse to be avoided, the plywood will flex, maybe/most likely enough to cause deflection of the panel with the possibility of developing cell cracks. Now you've got a bad micro/optimizer and a busted/compromised panel with cell cracks to boot and a nice story to concoct to avoid having a warranty claim denied.
Besides, unless there's space between panels which I'm (perhaps incorrectly) guessing the OP hadn't planned on if for no other reason than most folks don't, all that plywood rodeo, including at least tying it off to avoid sliding downslope, and also making sure the wind doesn't get at it ("40 ft. off the ground... and gets plenty of wind") and doing damage from impact somewhere - maybe even to the array, getting at the micros/optimizers will still probably require removal/replacement of a panel other than the one/several the plywood is on.

[bob-n]

JPM is right in every regard, as always.

I wasn't encouraging you to use plywood to work on panels, but it has been done successfully. If the plywood is thick enough and rests on the perimeter frame, the risk is low. I've even heard of people putting their full body weight on panels. I don't encourage that either.

String inverter, string+optimizers, or microinverter. String is simplest and puts the least electronics on the roof. Microinverters give you the best reporting and shade tolerance. APsystems can actually beat string inverters in cost, but perhaps not enough to matter. Unfortunately, we don't have enough failure data on inverters to give you confidence in any approach. All brands and all systems have had their issues in the past. We hope that they have learned and matured. Modern products seem to be doing pretty well. The only data I've seen says that panel failure rate is higher than microinverter or inverter failure rate. It's your choice.

As to 2/12 pitch, if that's what you have, you can still use it. Steeper would be better for all the reasons JPM gives, but it's really hard to tilt a building after it's built, and standoff racking is much more hassle that will not pay back quickly. I just ran a quick model of 9 degrees and 30 degrees tilt for an array in Dallas, TX. The 30 degree (7/12) array produced 6.5% more energy per year than the 9 degree (2/12) array. If my math is right, it's not enough to worry about. The dust accumulation is a bigger concern.

Any chance you can install a sprinkler system up there to wash off the panels every month?

High temperatures are bad for two reasons. Panels are less efficient as they get hotter, roughly -0.35% per degree C. And things don't last as long when they get hotter. That includes panels, electronics, and all of the rubber and plastic keeping things together. We're all at the mercy of the engineers that designed these things, hoping that they have enough margin. But again, you're not going to move the building to a cooler climate. The only thing you can do is allow air under the panels for a bit of cooling by using a racking system. I haven't seen calculations and don't know how much cooler that will make the panels, but laying the panels near the roof is clearly worse.

[J.P.M.]

I haven't seen calculations and don't know how much cooler that will make the panels, but laying the panels near the roof is clearly worse.

There's an NREL paper someplace that claims NOCT for a 5 cm. clearance is ~ 4 C higher than for a 15 cm. clearance. I did some cursory examination for arrays parallel to a sloped surface with less than a 45 deg. angle w/ the horizontal using some stuff from boundary layer theory and some Reynolds analogy stuff and came to a wild stab number of ~ 1 C increase in cell temp. for every 2.5 cm reduction in roof to bottom of panel gap up to a wind velocity of ~ < 3 m/sec. or so. I couldn't verify w/my array as I've got ~ 23 - 28 cm. clearance depending on how/where it's measured.

There may be and probably is other stuff in the open literature that I haven't seen/found/looked for yet.

Given the paucity of data on the subject as well as the variety of field conditions, I suspect the best to do until something more definitive and verified shows up is not try to quantify cell temp. as f(under panel clearance) too much but know that more clearance usually means lower cell temps. than less clearance, and so more annual production per STC W.

More clearance may also help keep the roof under an array dryer as well as (thinking long term maybe) aiding inspections for roof conditions.

[analogue]

Dust buildup can be a problem, particularly with low slope arrays, but in ways additional to those I'd thought of before installing my system. Most dust blows off or tends to accumulate in specific areas of panels and/or in a macro sense on sections of arrays, lower slopes --->>> more dust buildup. I've also found a bigger source of semi permanent fouling to be the dew that condenses on an array and turns the dust into caked mud. It doesn't blow away. It tends to build up. Low slope panels tend to turn into sandboxes, and the greater accumulation of dust, even though small, makes the caking worse for lower slopes as the mud runs down the array less. I note the bathtub ring on the lower 6"-12 " of many arrays and also note the generally lower width of the layer for higher sloped panels and vice versa.

FWIW, from what I think I might know, I'd be concerned or at least make some provisions for easier access to the array to be able to hose it off from the top 1X in awhile.

Also, w/no shading why are you using optimizers or microinverters ? More $$ or at least more to go wrong than w/a simple string inverter. The KISS principle applies. The higher probability of required service as system complexity increases is also another reason to consider easy access.

IF it was my project, I'd also run my plans past the AHJ before I bought anything and to check with them for any other requirements they may have (if any) such as for external loadings such as those incurred from wind or seismic that I might not be aware of. I sure don't know about your local codes, but I'd also check for required setbacks for first responder access. That array looks a bit tight. Also, if you use optimizers or micros, do you intend access to the panels in the middle of the array when service is required ? If so, how ?

Lastly, maybe I'm wrong, but it looks like the rail free systems you post don't allow much free space for ventilation under the array. That'll mean increased cell temps. and so lower system efficiency and so lower production.

Welcome to the neighborhood.

Take what you want of the above. Scrap the rest.

Your reasoning for dust build up does make a lot of sense. Especially the morning moisture/dew cycles leading to the caking of whatever the heck happens to be on the panels at the time. Can't say that I'm too excited about going up there to hose things off too often, but bob-n's idea about installing a sprinkler is something I'm crazy enough to consider doing. Maybe even hook up a security camera to be able to keep an eye on things if I'm bored enough

Now, on to the juicy stuff - microinverters vs optimizers vs string inverters. I'm a bit of a data nerd so I'd initially set a hard requirement that panel level monitoring is a must have regardless of conditions. This was before I'd done any research into the myriad of concerns regarding one technology vs. the other. For this reason, I short-listed Enphase and APSystems microinverters. I'm also a bit opinionated regarding local data access, data ownership, privacy, and dependence on the cloud. APSystems is pretty terrible in all these regards and they are not exactly DIY installer friendly (there goes the warranty) so they're a last resort. Enphase is considerably more expensive (2x APSystems) but that is a tradeoff I might have to make.

So just over the last few days, I started looking more into the heat related concerns for microinverters under the panels. I had discounted it previously based on all the guidance (or propaganda as some might call it) out there, but the deeper I dug, the more my confidence wavered. Specifically, every single thing I found on record from manufacturers and installers that manufacturers found to go on the record worded things in a very specific way. The phrase "we've never had a microinverter system failure due to heat" is bandied about quite a bit, but no one ever addresses the gray area - specifically, no one talks about the degraded performance due to heat. There is enough reading in between the lines there that one could assume manufacturers are very tightly controlling the narrative to make sure focus is not put on performance. Couple that with the brutal Texas summers and the fact that it is only going to get worse over time, microinverters don't sound like the best rational choice. I did find this quote from an installer in the Austin area. Basically, heat related issues due to microinverters have caused them to drop microinverters altogether.

So where does that leave things? Well, I'm seriously considering plain old string inverters (no monitoring) or power optimizers (for the monitoring, not the optimization). I'm of two minds at the moment and things seem to have landed on two manufactuers - SolarEdge and SMA/Tigo.

Just based on initial research, I'm already not a fan of SolarEdge + their power optimizers. Data wise, they're locking things down to the cloud, not allowing local access, and it seems part of their whole system reeks of proprietary tech in the optimizer (nothing based on SunSpec) so if/when I need to replace things in the future, I'm at the mercy of SolarEdge. I didn't spend any more time on them after getting soured on their data access stance. For those that are curious, my methodology isn't exactly scientific, but being able to find projects on Github that allow local data access is a pretty good indication. For those with hands-on experience with SolarEdge, is my assessment accurate?

The last option is plain old string inverters with SMA. One thing I did learn that I wasn't aware of before is the whole NEC 2017 Rapid Shutdown requirement. It seems like the back of every panel is now going to need an addon electronic module that conforms to SunSpec Alliance protocol for rapid shutdown. Sounds reasonable and is not too expensive ($30-$40/panel) but then there is also the option to get an optimizer coupled with it for not too more ($60/panel total) that allows per-panel monitoring. These are both Tigo modules, not SMA. It seems like we've come full circle though, back to having an electronic module under the panel to get abused by the Texas heat. I can't tell if the SMA -> Tigo comms for monitoring are proprietary or not, but as you probably know by now, I'm not a big fan of proprietary systems and prefer open spec based comms and protocols. Are my conclusions accurate here for folks that have experience with SMA/Tigo systems?

Looping back to the reason I posted this topic, it is quite apparent that a rail-free system reduces the amount of space under the panel and that can only make heat related concerns worse. In fact, this was actually pretty useful because I now know not to go with "light" rails which are not as high off the surface of the roof as "standard" rails.

When folks talk about AHJ, are they usually talking about their local electric utility or some other governing entity? I'm on PEC (Pedernales Electric in the Austin area) and they seem pretty hands off about the whole thing other than inspections and the like. The only other constraints I know of are my HOA which mandates that panels are installed parallel to the roof line (no additional tilt allowed for optimization).

You're spot on when you say the array looks a bit tight. I happened to stumble across PVSketch and modeled my array. Turns out I'm going to have to settle for an array more in the 15-17kW range instead of the net zero 20kW array that I had my eye on

Really appreciate your thoughts and looking forward to learning more!

[analogue]

JPM is right in every regard, as always.

I wasn't encouraging you to use plywood to work on panels, but it has been done successfully. If the plywood is thick enough and rests on the perimeter frame, the risk is low. I've even heard of people putting their full body weight on panels. I don't encourage that either.

String inverter, string+optimizers, or microinverter. String is simplest and puts the least electronics on the roof. Microinverters give you the best reporting and shade tolerance. APsystems can actually beat string inverters in cost, but perhaps not enough to matter. Unfortunately, we don't have enough failure data on inverters to give you confidence in any approach. All brands and all systems have had their issues in the past. We hope that they have learned and matured. Modern products seem to be doing pretty well. The only data I've seen says that panel failure rate is higher than microinverter or inverter failure rate. It's your choice.

As to 2/12 pitch, if that's what you have, you can still use it. Steeper would be better for all the reasons JPM gives, but it's really hard to tilt a building after it's built, and standoff racking is much more hassle that will not pay back quickly. I just ran a quick model of 9 degrees and 30 degrees tilt for an array in Dallas, TX. The 30 degree (7/12) array produced 6.5% more energy per year than the 9 degree (2/12) array. If my math is right, it's not enough to worry about. The dust accumulation is a bigger concern.

Any chance you can install a sprinkler system up there to wash off the panels every month?

High temperatures are bad for two reasons. Panels are less efficient as they get hotter, roughly -0.35% per degree C. And things don't last as long when they get hotter. That includes panels, electronics, and all of the rubber and plastic keeping things together. We're all at the mercy of the engineers that designed these things, hoping that they have enough margin. But again, you're not going to move the building to a cooler climate. The only thing you can do is allow air under the panels for a bit of cooling by using a racking system. I haven't seen calculations and don't know how much cooler that will make the panels, but laying the panels near the roof is clearly worse.

The gears are seriously turning on the sprinkler system idea. No electricity necessary and I just have to figure out a way to mount the sprinkler doohickey to the roof and how to snake a hose up there that will have enough water pressure left to actually do some sprinkling!

See my response to J.P.M on the current state of things wrt dropping microinverters altogether. Also, thanks for the ideas and input - all good stuff!

[foo1bar]

Just based on initial research, I'm already not a fan of SolarEdge + their power optimizers. Data wise, they're locking things down to the cloud, not allowing local access, and it seems part of their whole system reeks of proprietary tech in the optimizer (nothing based on SunSpec) so if/when I need to replace things in the future, I'm at the mercy of SolarEdge.

Data is "in the cloud" - but is also available for download. And it seems like they've played nicely with websites like pvoutput.org

There is a risk that if an optimizer breaks and if solaredge is out of business and if no one has bought the tech during bankruptcy that you could be down a panel until you redo the inverter setup.
I decided I was willing to take that risk - my belief is/was they're likely to survive at least 5 more years. And if they do go bankrupt at some point in the future IMO it's likely someone would buy enough of the assets to provide parts/support for the preinstalled base.
And I had more belief in them and their approach vs. enphase which was the other possible option I had.

When folks talk about AHJ, are they usually talking about their local electric utility or some other governing entity?

AHJ = Authority Having Jurisdiction.

Usually the city (or county) building inspection department.

You're spot on when you say the array looks a bit tight. I happened to stumble across PVSketch and modeled my array. Turns out I'm going to have to settle for an array more in the 15-17kW range instead of the net zero 20kW array that I had my eye on

15-20kW array in central texas?
Seems really large to me. You'll want to make sure you won't have any issues with >10kW system (10kW can have different rules with the PoCo and/or the AHJ. Also electrical service / main panel needs to be able to handle that much power.
And I'd double check what the real need is. are you running a pool pump a lot? Or some other motor? Can that load be reduced?

If you really want/need 10kW, possibly you can still do it with higher watt/square-meter panels.
But that'll cost you more of course, so it may be better financially to not do that.

Also- I don't know the rules where you are, but in some places there are requirements for setback from the edges and ridgeline of the roof. The idea is to allow firemen to be able to access the roof easily to cut holes, etc when fighting a house fire. So before you get too far along with the planning of how you'll have your array, check on that. And some places that have those rules allow the local fire department to give waivers (My AHJ allowed my fire dept to say mine was OK, even through I don't have the required 3' setback)

[analogue]

Data is "in the cloud" - but is also available for download. And it seems like they've played nicely with websites like pvoutput.org

There is a risk that if an optimizer breaks and if solaredge is out of business and if no one has bought the tech during bankruptcy that you could be down a panel until you redo the inverter setup.
I decided I was willing to take that risk - my belief is/was they're likely to survive at least 5 more years. And if they do go bankrupt at some point in the future IMO it's likely someone would buy enough of the assets to provide parts/support for the preinstalled base.
And I had more belief in them and their approach vs. enphase which was the other possible option I had.

Did some more research and SolarEdge is back in play. There seem to be workarounds to get local access but it is going to be a game of cat and mouse (firmware update closes exploit, new exploit is found, rinse and repeat). Turns out Tigo actually charges a subscription for any useful data so they're out unless I give up on per-panel monitoring just use them for rapid shutdown compliance.

AHJ = Authority Having Jurisdiction.

Usually the city (or county) building inspection department.

I was able to track down more information on this too. Turns out the county fire marshall is the one to go to and they defer quite a bit to the 2015 International Fire Code (IFC). There are even special provisions for 2/12 pitch roofs where the setback requirements are pretty much non-existent (page 83).

15-20kW array in central texas?
Seems really large to me. You'll want to make sure you won't have any issues with >10kW system (10kW can have different rules with the PoCo and/or the AHJ. Also electrical service / main panel needs to be able to handle that much power.
And I'd double check what the real need is. are you running a pool pump a lot? Or some other motor? Can that load be reduced?

If you really want/need 10kW, possibly you can still do it with higher watt/square-meter panels.
But that'll cost you more of course, so it may be better financially to not do that.

Our electricity utility has recently expanded their max size for residential systems from 20kW to 50kW so I'm pretty happy about that. I am taking into account degradation over time, higher load in the future (an electric car or two), non-optimal tilt 10 degrees), and we're 100% electric (no gas to speak of) for all household loads.

Also- I don't know the rules where you are, but in some places there are requirements for setback from the edges and ridgeline of the roof. The idea is to allow firemen to be able to access the roof easily to cut holes, etc when fighting a house fire. So before you get too far along with the planning of how you'll have your array, check on that. And some places that have those rules allow the local fire department to give waivers (My AHJ allowed my fire dept to say mine was OK, even through I don't have the required 3' setback)

I lucked out on a 2/12 pitch roof (well, I actually had it as 1/12 during design and it got bumped to 2/12 for standing seam metal warranty eligibility). Supposedly I can build right out to the edge of the roof but I'm going to try to keep it reasonable for service access when needed.

Thanks for the input!

[solarix]

The Tigo RSD modules are a relatively simple switch to turn off the panel in emergencies and as such if they fail, they tend to leave the panel working. However, when microinverters or Solaredge optimizers fail, you get zilch form the panel.