Skip to content

derek the solarboi Posts

“Racing” Toward Perovskites – TL;DR

From CleanTechnica:

In a new paper published February 26 in the journal Nature Energy, a CU Boulder researcher and his international collaborators unveiled an innovative method to manufacture the new solar cells, known as perovskite cells, an achievement critical for the commercialization of what many consider the next generation of solar technology.

Here’s the TL;DR – No, perovskites are not dramatically closer to being commercially viable. They found they could reduce oxidation in open air (which is what kills the performance of perovskites over time) by mixing dimethylammonium formate with the perovskite solution before it’s sprayed onto the panel. This allows for retained performance of 90% after 700 hours, up from 300 hours. As the article notes, there are over 8000 hrs in a year.

Still have a ways to go.

Tesla Roof comparisons by Matt Ferrell

Absolutely fantastic video from Matt Ferrell here. He does a pretty fair apples-to-apples comparison between his more typical solar setup and Paul Braren from‘s Tesla solar roof installation. Where I often see very abstract cost comparisons to Tesla’s solar roof, this takes a look at the costs and benefits in a ton of different ways. Both systems are definitely higher in cost than your typical solar install, but the comparisons are fair and useful! Loved this.

Fiberglass PV Frames?!

From pv magazine:

“This composite material is used in applications such as wind turbine blades, to withstand wind pressure, vibration and centrifugal force, as well as railway tracks, to withstand the pressure and vibration of passing trains,” a spokesperson from the company told pv magazine. “Fiberglass-reinforced composite materials have been used for over 20 years in outdoor environments and fields with higher load requirements, with countless successful application cases.”

Fiberglass-reinforced composite for panel frames is fascinating on a number of levels.

  • Panel frame wouldn’t need grounded, which is a very weird thing to think about.
  • Would it make the panel heavier or lighter than aluminum frames? Presumably heavier.
  • Not having to deal with any leakage to ground through the panel frame in 20 years might actually lead to better longevity.
  • Then again, I don’t know how fiberglass-reinforced composite will fare after 30-40 years in direct sun. We know the silicon lasts forever if it’s treated well, but will the composite match aluminum’s durability?
  • This would be worse for recycling

What do you think?

SMA Core 1 inverter loses serial number!

Ever had an issue with SMA inverters losing their serial number? Here’s how to fix it!

This is an example of this happening to me recently, and I’ll give some more details below the video, if you’re struggling with this same issue.

I’ve only ever seen this happen on the first iteration of the Core 1 inverter, the -40 model. The symptom is a wifi signal from the inverter that’s showing a “0900” number in the SSID. For instance, in a normal SSID, I’d expect to see the name be “SMA3000XXXXXX”— basically, “SMA”, plus the serial number visible on the exterior inverter label. But when the inverter loses its serial, it makes up a completely different serial for itself, usually starting in “0900”. In my case, I was seeing “SMA0900022963”.

This issue can cause problems with communication and general administration of the device, so it should be fixed, if possible. However, I’ve never seen this issue be responsible for lost power, so it’s a low-level problem, in the grand scheme of things.

The fix for this can only be accomplished through Level 2 of SMA’s tech support (call +1 (877) 697-6283). Level 1 will get your information, tell you your case number, and send you to Level 2.

Level 2 will need to screen-share with your laptop (Windows only), while your laptop is on the same network with the inverters themselves. You need to visit the login page of the inverter, but instead of the ending url being “login” (i.e. “https://<IP-Address-of-inverter>/#/login”), it needs to be “loginadvanced” (i.e. “https://<IP-Address-of-inverter>/#/loginadvanced”).

From there, to log into the “Service” user group, SMA’s tech will need to generate a password from that serial number you saw in the WiFi SSID earlier. In my case, “900022963”. There’s often a lot of confusion from the technician at this part. They’ll be fine, they’ll figure it out eventually. Last I knew, their password generates from the “serial number”, WITHOUT any zeros at the beginning.

Once the Service user group has been logged into, the only thing that needs done is typing in the correct serial number in Parameters->Type Label->Type Label->Serial Number.

Easy peasy, but you’re heavily reliant on that password from SMA for this fix.

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.

Solar Adoption in Rural America

From Mike De Socio at CNET:

“We were able to find that adoption of these technologies is highly related to income,” Mayfield said. “We also find that education is also a main factor of these technologies.” In other words, rural Americans with higher incomes and more education are more likely to put solar panels on their roof or buy a heat pump.

I mean, yep.

And the big “solution” many companies turn to for low-income and low-education folks, is leasing. I’ve met maybe a handful of people who were a big fan of their lease, but they were markedly at the beginning of the craze in the early 2010s when the deals were much better. However, most leaseholders I’ve encountered, especially in recent times, have become cynical about solar, mainly due to poor service.

Solar leasing companies historically have very few service people available, and it often takes months of hassle to get people to come out. The usual scenario involves residents reaching out to the leasing company for months without a response or with dismissive gestures. As a last resort, the resident stops paying the monthly bill, finally grabbing the attention of the company. However, instead of getting someone to fix the system, the company contacts someone like me to disable the system outright until payment continues.

Every single one of those kinds of jobs I’ve seen are on low-income housing. When these leasing companies mess up, it reinforces the belief that solar is a scam. Paired with fast-talking salesmen with no morals, leasing is the fastest way to introduce solar to low-income Americans — and a recipe for alienation.

Best way to educate, as with anything you want to incentivize, is to increase the accessibility of ownership.