Skip to content

Tag: solar

Straight to Jail – RE+ 2024 Hall of Shame

I get press missives now, and I guess Mercom Capital sends the press a “top 10 startups” list after RE+ every year. I had fun looking through it – there are some eye-rolls, as there always are when big-money folk try to assess what technology has a meaningful chance to make it big, but one, in particular, stood out to me.

Ladies and gentlefolk, may I present Sundial, a supposed universal tracker.

Sundial is billed as three things:

  1. A luxury upsell, to increase the installer’s revenue per project
  2. Allow for selling smaller systems, reducing cost to get more customers through the door
  3. Offer retrofits for customers who are unhappy with the performance of a system they were sold

On the plus side, I’m sure getting parts for this thing if/when the moving parts fail would not be difficult, because it looks like something built in someone’s garage. All the promo videos appear to be sped up, and the motor is extremely loud.

Considering this would be ADDED to a “normal” installation (fastened to whatever racking system you have, with the module attached to the Sundial), you need one for every single panel in an installation. Imagine 20 of those motors on your roof, whining throughout the day.

It also looks extremely inflexible. It says it can be for seasonal tracking or daily tracking, depending on orientation, which is laughable to me. For instance, on a residential system with a few rows of modules in an array, just about any individual panel tilt is going to end up shading another module unless you design the system with farther spacing. The tile angle only goes in one direction, starting with the face of the roof. There are very few situations where this might have a meaningful impact on production, which all but eliminates selling point number 3.

That’s not even the worst part about trying to sell this to dissatisfied solar customers. To install this, you’re adding extra weight, and you’d need to redo engineering/permits. You’re adding greater system complexity with motors. If God forbid, one of the 10-20 motors you installed on someone’s roof fails, and they’re ALREADY unhappy? That’s a nightmare for customer experience.

As for selling points 1 and 2, number 1 is a despicable reason to pick up a product by itself. Luxury upsells are fine, but you need to be talking about how they bring the customer value. For number 2, I cannot see a way that this will result in less cost overall for customers unless they have an incredibly good price point and an amazing ease-of-installation story.

To be frank, I do think it’s a fun idea. But maybe just for hobbyists. It’s not a product that has real potential in the larger market. Its failure will be from trying to accommodate far too many situations without an extremely good reason to exist.

Rocking Solar, another company featured in Mercom’s startup list, is an interesting contrast here. Their scope is extremely narrow, and they do a lot of work to emphasize the reliability of the parts they’re using. They aren’t for retrofit. They’ve narrowed the motor/panel ratio to potentially 1:100. It’s still added complexity and a bunch of moving parts (which by default makes me grumpy), but they have numbers to support the product and specific, sensible reasons why you might use it.

Ultimately, I’m ranting because the big-money people often don’t understand the products they give attention to. I feel like Sundial either had excellent connections to get this sort of attention, or someone at Mercon said, “oh cool, a residential tracker, I haven’t seen that before, so it’s gotta go on our list.”

It’s a fun demo but a stupid product, as it stands right now.

Pegasus suggests MASSIVE changes for UL 3741

Editor’s note 07/09/24 – An earlier version of this article asserted that any module could be used with the Pegasus railing. That was incorrect, and the offending intern has been sacked.

Amid Unirac publishing their new listings for NXT Umount and SolarMount, and rumor of EcoFasten getting their railings listed, Pegasus may have just completely blown the roof off the expectations for 3741 listings.

To date, the onus of testing has been placed on the racking manufacturer, who gathers all the ingredients, in dozens of permutations, to test components like modules, inverters, and MLPE. All those components are then specifically listed in the racking manufacturer’s “3741 installation addendum”, or some other document that modifies the installation manual for 3741 applications.

However, Pegasus contacted me this week to show me their listing: one with ZERO approved inverter vendors lists. No inverters, no MLPE. Excuse me, but what in the actual fuck.

I initially assumed that there was a document they forgot to send, so of course I asked for a list of approved vendors. They wrote back that no, there was no mistake. They simply don’t have a list of approved vendors, and that as long as the equipment used is listed under UL 1741 (or has some other specific UL 3741 COC), you may use that equipment with their railing.

ANY inverter. Not just an inverter that’s been tested by some other racking manufacturer, but ANY inverter. The implication is that literally everyone else could do this, too. We could find ourselves in a situation where picking an inverter and rail system that works with UL 3741 will no longer feel like untying a Gordian knot; instead, we’ll be slicing through it with a sword.


I have so many feelings about this— a mix of excitement, frustration, and dread. I’m excited that this indicates the industry could move to a much simpler way of listing these products, spending less of humanity’s time on confusion, lowering the barrier to entry, and allowing for more solar that is less complex. I’m frustrated that only now, 2 years after the first listing with SMA and Sollega, have we discovered that this is possible. And I’m dreading that the code body will see this, say “Oh, you’re not supposed to be able to do that”, and change the standard to lock out this option.

There’s probably good reasons why racking companies have defaulted to listing every piece of third party equipment, but it seems to be an overly large burden. It also ends up being far more complicated for the end user than it should be. Even if Pegasus’ strategy isn’t technically correct in some way, I hope we can compensate by adding the appropriate requirements into inverter listings, instead of overstuffing racking listings with redundancy.

Truly, the biggest obstacle to greater UL 3741 implementation is confusion about what equipment can be used (and also SolarEdge/Bill Brooks trying to undercut it and scare everyone with any chance they get). This is the entire reason I created UL3741.com, and dear reader, I truly hope this action by Pegasus will lead to the website’s demise. Simplifying the listing like Pegasus has done is the coolest thing ever. I just hope that we get to keep it.

Unirac is NXT to Residential UL 3741

Intertek has posted the updated Photovoltaic Hazard Control listings for Unirac on their directory, and it includes a few new racking models.

Included in the listing are existing racking models, which are exclusive to flat roof solutions, but newly added to the list are NXT Umount and SolarMount, with callouts for both residential and commercial.

The full list of new addendums:

  • UL 3741 PV Hazard Control Installation Addendum for NXT Umount Residential Roof Applications PUB2024MAY07
  • UL 3741 PV Hazard Control Installation Addendum for NXT Umount Commercial Roof Applications PUB2024APR26
  • UL 3741 PV Hazard Control Installation Addendum for SolarMount Residential Roof Applications PUB2024MAY07
  • UL 3741 PV Hazard Control Installation Addendum for SolarMount Commercial Roof Applications PUB2024APR26

We don’t know yet what inverters will be listed with these racking systems, or exactly what installation will look like, but it’s safe to assume that we can expect something similar to Ironridge and Chiko’s listings.

Chiko is the only company so far to have listings for a sloped-roof racking that includes commercial inverters, but it sounds like Unirac will be next to the party. This will be immensely helpful for solutions like agriculture, where there’s often large, accessible, sloped roofs that could accommodate roof-mounted inverters.

This is as yet unannounced by Unirac, so the cited documentation in the listing is not available publicly. I imagine we’ll see more details in the coming weeks. Update: Unirac has now officially announced these listings.

If you want to know the details of the yet-to-be-released documents once they’re available, make sure to subscribe to the newsletter at UL3741.com. While you’re there, you can explore all current UL 3741 listings, cross-referenced with compatible inverters and array-level rapid-shutdown systems.

Crystalline Solar is Forever

Editors note: When initially publishing this, I was unaware that the sample size for the study in France was pretty small and there may be some notable omissions from the PV Mag article. Keep your brain in gear, dear reader.

Perovskites punching air rn.

From Gwénaëlle Deboutte at PV Magazine:

New testing conducted at France’s oldest PV system have shown that its solar modules can still provide performance values in line with what the manufacturers promised.

The testing showed that the modules still produce on average 79.5% of their initial power after 31 years of operation. In a previous testing carried out 11 years ago, the panels were found to produce 91.7% of their initial power.

An average of 0.66% degradation every year over the course of 30 years of use is incredible, even though it’s clear that the third decade brings more degradation than the first two. The fact that solar panels I installed 10 years ago will still have significant usefulness in 20 years is amazing.

It IS likely that most solar panels won’t be given the option to last for 30+ years in their original place of installation. Investors are usually gonna scrap the slower-replicating money in chase of the faster-replicating money, as technology advances. However, I wonder if we’ll see an increase in selling these older panels to countries that aren’t as desperate to get the latest solar tech. After all, the best alternative to recycling is reuse.

I’ve been convinced that perovskites, what we’re all looking toward as the proverbial future rocket fuel in this industry, could never dream for this kind of reliability. Reliability has long been the kryptonite of this technology. It’s a problem that we keep hearing is on the verge of a solution, as this study from Rice University claims:

A Rice University study featured on the cover of Science describes a way to synthesize formamidinium lead iodide (FAPbI3)—the type of crystal currently used to make the highest-efficiency perovskite solar cells—into ultrastable, high-quality photovoltaic films. The overall efficiency of the resulting FAPbI3solar cells decreased by less than 3% over more than 1,000 hours of operation at temperatures of 85 degrees Celsius (185 Fahrenheit).

“Right now, we think that this is state of the art in terms of stability,” said Rice engineer Aditya Mohite, whose lab has achieved progressive improvements in the perovskites’ durability and performance over the past several years. “Perovskite solar cells have the potential to revolutionize energy production, but achieving long-duration stability has been a significant challenge.”

This is, let me be clear, WAY better than anything we’ve seen so far. I don’t know how comparable 1,000 hrs of operation at 85°C is to one year in an Arizona desert, but that’s way more successful than anyone else has been so far. Just this February, advancements were being trumpeted about only 10% degradation after 700 hrs of operation. But given that there’s about 8,000 hrs of sunlight a year (depending on a number of factors, obviously), perovskites are going to have a hard time reaching that <1%-degradation-per-year threshold.

However, some companies are already convinced they’ve fixed the degradation issue and that perovskites are ready for prime time, as Marian Willuhn writes in PV Magazine:

Utmo Light, a Chinese perovskite company, is showcasing its first commercial perovskite solar module this week at the SNEC PV trade show in Shanghai. The Module UL-M12-G1 measures 1,200 mm x 600 mm and is available in four power classes, ranging from 110 W to 130 W.

Utmo Light President Zhenrui Yu told pv magazine at SNEC in Shanghai that the first target for the new perovskite modules will be building-integrated PV (BIPV) applications.

Yu said the panels have passed all IEC testing for solar modules and can withstand a 2,300-hour UV bath at 1,000 watts per square meter and 60 C, for 12 years of operation without degradation.

These panels are 1200m x 600mm, with a max of 130w, with a cost of around $.19 per watt (not accounting for whatever tariffs may or may not apply). That IS better than the cost of roughly $.30 per watt for a crystalline panel. But the max efficiency you can get from these perovskites at this stage of production is about 18.06% (180.6 w/m² at 1000 w/m² of irradiance).

To take a random contrasting example, Axitec’s 550w monocrystalline panel (AC-550MBT/144V) is 2278mm x 1134mm, making a total efficiency of about 21.29% (212.9 w/m² at 1000 w/m² of irradiance).

So comparatively, this commercial perovskite panel from Utmo is extremely meh as an introductory product, given its lackluster efficiency. Despite a (maybe?) better price, it can’t currently touch crystalline silicon in production.

This does make sense, though. As the Rice University article mentions, the higher efficiency perovskites are the most unstable, so Utmo has probably solved for 2D perovskites (most stable, least efficient) and plans to build up from there. 3D perovskites, which contain the power-density we all dream of, still have a ways to go before they hit the market.

There’s a part of me that is hopelessly skeptical of how much time and effort it has taken to get this technology to market. How is it possible that this can EVER be stable, when it feels like all the resources of the planet can’t get us to where we want to be, after decades of research?

Yet at the same time, I didn’t have a front row seat to the development of the ever-lasting crystalline tech that our industry now depends upon. I didn’t have the chance to watch research develop, or companies hype unfinished technology, so it’s easy to take for granted the struggles of the past while I criticize the struggles of today.

I will continue to be skeptical, but perhaps…crystalline solar may not be forever.

Shining a Light on Lumio

From Eric Peterson at Utah Investigative Journalism Project:

McDole has hired a lawyer and is still battling the company over her claim of roof damage and for giving her what she said is false information about tax rebates being able to offset a third of the $40,000 loan she had to take out.

When McDole heard about Lumio’s tax incentive, she was astounded.

“Where’s my fucking tax break?” she asked.

Sigh. Another day, another large solar company accused of defrauding their customers. The entire article’s very good, go read it and then come back.

I keep getting Pink Energy vibes, here. For better or worse, I’ve been obsessed with the Powerhome Solar/Pink Energy story recently, along with diving into the CEO’s claims and actions over the course of the last several years. Pink Energy, of course, went bankrupt in 2022, after years of customer complaints— similar to those being brought to Lumio. Of course, Pink Energy’s death was brought much more swiftly due to issues with Generac’s solar product at the time, but I tend to think Pink Energy wasn’t going to last, anyway. I’ve seen enough reports of bad training and mismanagement at the lower levels that is really hard to fix, especially when the CEO bragged in his own book about firing 20% of Powerhome staff, every time the performance numbers weren’t where they needed to be1.

Solar is complex. Honestly, more complex than it often needs to be, mostly due to regulatory shit. Changing codes, especially when you’re a company that’s multi-state, can be a nightmare for training designers and installers. And I don’t know for sure what Lumio’s business model is, either, because there’s a ton of conflicting information about whether Lumio’s installer’s are all in-house or if they use a number of contractors (I dare you to try deriving any conclusion from this Reddit thread, for example). Regardless, they started by buying or merging with a number of solar companies. Especially in that situation, quality and customer care becomes a super difficult thing to manage and maintain under the expectation of immediate growth, a problem with most investor-backed companies.

Growth at all costs is extremely costly. And in this industry, growth at all costs for a solar company usually ends up being more costly for the customer. Large solar companies are often able to hide behind well-paid lawyers, flashy marketing, and tedious games of phone tag, where the solar company never tags back.

And again, we come back to the question: Will solar ever get to a place where we’re not identified largely as scammers, and if so, what are the market forces that will finally make it happen? Do we need yet more regulation, this time in sales contracts? Do we need “Nutrition Facts” labels for solar contracts, explaining in consistent terms how if you don’t have the tax liability to get your 30% credit, you’re gonna suddenly be paying way higher monthly bills for the next 20 years? Do we need some faster, state-enforced means of forcing lazy solar installers to come back and fix their fucked installations? Or is this simply an inevitable by-product of the opportunity presented by federal tax credits and grants, seized by salesmen and CEOs with dollar signs in their eyes?

Just stop screwing up my industry, geeeeeez

  1. Own Your Power by Jayson Waller, p138 ↩︎

Issue Solved: Fronius MPPT1 dropping production

I’ve been fighting with two identically installed Fronius Symo Advanced (10kW) inverters over the last month, and I think I’ve finally fixed it. It’s very strange, and seems to be an issue in the firmware of the inverter, itself.

The installation looks like this. They are two carports (only one is pictured), identically oriented, with north, south, east, and west arrays. The east and west arrays are smaller and hooked up to a Fronius Primo 3.8 on each carport, obviously using different MPPTs for different arrays. The same thing is the case for the north and south arrays: they’re hooked up to a Symo Advanced 10.0 at each carport.

The problem is that MPPT1, hooked up to the north array on each carport, drops off during high irradiance days. Check it out.

You can see when it starts producing in the morning, drops to nothing, then later in the day, it jumps back up and starts producing. Sometimes it takes until after noon, and sometimes, like pictured above, it comes back up when the irradiance dips and MPPT2 produces less for a bit.

This only happens on higher irradiance days with good, bright sun, and on both carports. On cloudier days, both north and south arrays on both carports produce equally, as you’d expect with diffuse sun.

Fronius tech support was no help with this one. We made sure firmware was up to date (fro34310 at the time, for those following along), but beyond that they were extremely confident that there was something wrong with our wiring or our array setup, but couldn’t give me proper direction. It may have simply been the one tech I was talking to, but he was extremely frustratingly not budging on his assessment. It doesn’t help that there are no error codes triggered with this problem.

In troubleshooting, I couldn’t nail any issues to our set up for rapid shutdown devices; everything on site is rated to work with each other and is set up properly. Had some previous issues with dueling RSD transmitters tripping arc faults that I got sorted, so very confident that’s not an issue any more.

When I was onsite, however, I was able to shut off DC to the inverter and turn it back on within about 10-20 seconds, and the inverter jumped up to the proper production on all strings, so it would seem there’s nothing instantaneous in the array that’s causing MPPT strangeness.

Finally, I swapped MPPTs between arrays, putting the south array (highest producing) on MPPT1 and the north array (lowest producing) on MPPT2. After giving it some time, I think this one change solved my issue.

You can see in comparison to the first graph, that the north arrays seem to have no more issues with dropping off in production. The south arrays didn’t have the irradiance on this day that they had in the previous graph, but it’s enough that the north arrays would have dropped off if they were wired as they were originally.

My reckon is that this is directly related to the inverter wigging out when MPPT2 gets 5x or more the production of MPPT1, and only resets when MPPT2 stops rising. I’d imagine that most installations would put the main and highest producing arrays on MPPT1 by default anyway, so this issue would be very rarely seen. If you’re having this issue, try making sure MPPT1 gets the most production between the two, and it might fix it for you.

Weird weird issue.