microinverters vs optimizers

[Solar_OR]

I am wondering if anyone is seeing a shift towards one or the other these days? I figure Rapid Shutdown is going to cause for a push. I have used Solaredge and Enphase in the past. I had fairly good experience with both. And am curious to see how the new Enphase system rolls out and their battery. Does anyone have experience using APsystems? The initial cost seems to be cheaper than Enphase and I like the idea of installing fewer inverters. But I have never used them before so I am not sure on ease of use. Reliability? Customer service?

Any feedback is greatly appreciated. I would love to hear some real world considerations.

Thanks!

 

[SolarPro]

We decided this was worthy of a SolarPro article:

Integrator Perspectives on Module-Level Power Electronics

It is super interesting to see how much a Code changes can change the market for solar power electronics. While there are early adopters for MLPE in every market, as soon as the 2014 Code get adopted, companies that only used MLPE selectively, if at all, can pivot away from string inverters almost over night. Our 2015 article captured some of this. I saw it first hand here in Austin a few months later.

Back when we published this MLPE article, APSystems probably wasn't on a lot of people's radar. Their market share is definitely increasing. For the most part, it seems like SolarEdge has benefitted the most from rapid shutdown enforcement, at least in terms of a % market share increase.

 

[Solar_OR]

Thank you for the reply and article. I agree they seem to be gaining speed lately and some of the other microinverter companies exiting the game may have propelled them. The article points to a lot of the concerns I have had as far as reliability of the products as well as the company. 25 years is a lot of time. And the cost to troubleshoot/service the systems is certainly something to be taken into consideration. Extra points of failure and proprietary components in the system.

It has made sense in residential at times from my experience because of shading, various roof surfaces, and other factors. But, for commercial projects, I have not seen the need/benefit. String inverters on the roof has been fairly easy to implement. 2017 NEC will definitely make that a little more difficult to comply with 690.12 so that is why I am wondering about the feasibility of some of the microinverters out there today. Have you seen many 690.12 compliant commercial projects in a 2017 NEC jurisdiction? Plus the additional labor and materials for commercial MLPE systems seems tough.

 

[SolarPro]

Wouldn't you know it, there's also a SolarPro article for that:

Module-Level Rapid Shutdown for Commercial Applications

After researching this topic and speaking to stakeholders, I think it's far from a foregone conclusion that the UL's new rapid shutdown safety standard will ultimately require module-level rapid shutdown. MLPEs will certainly be one type of system that can comply with whatever consensus safety standard gets adopted, but I suspect that string-level products like the HiQ inverters will also be able to reduce hazard to an acceptable level. That may provide an opening for string inverters with multiple MPPT inputs, as that would seem to reduce available fault energies. I'm just guessing, but I'd be surprised if UL develops a standard that only allows for module-level shutdown.

Otherwise, I agree with your basic line of thinking. It's difficult to justify MLPE in commercial applications. Microinverter systems don't scale well. Sure, you can do what APSystems has done and make a multi-module microinverter, but your basic building blocks don't provide a lot of opportunity for cost reductions as systems get larger. The SolarEdge approach seems to be more interesting because it allows for very long source circuits, which does enhance the platform's value proposition. SolarEdge basically argues that you can get the advantages of 1,500 Vdc strings with it's commercial platform, based on its fixed input voltage; best of all, you can use standard 1,000 Vdc rated components, which sounds intriguing. The TenK platform is also pretty interesting once you wrap your head around it. But you have to evaluate the platform based on LCOE and be able to convince your finance partner that the equipment and production estimates are bankable.

Great technology is only a piece of the puzzle...

 

[pv_n00b]

If the stock market performance is any indication the favorite is MLPE over micro-inverters.

 

[Solar_OR]

Enphase has been in the red for awhile and the recent amount of layoffs may not be a good indicator...

 

[jaggedben]

To my eyes in the residential market rapid shutdown pretty much wipes out the cost advantage for string inverters, which are already compromised by design inflexibility. Commercial and residential markets are quite different, in multiple ways, that favor microinverters and optimizers in the latter more than the former.

I'm going to keep conclusions of higher level cost analysis closer to my chest. Can't give away all my secrets. :happyno:

 

[ggunn]

To my eyes in the residential market rapid shutdown pretty much wipes out the cost advantage for string inverters, which are already compromised by design inflexibility. Commercial and residential markets are quite different, in multiple ways, that favor microinverters and optimizers in the latter more than the former.

I'm going to keep conclusions of higher level cost analysis closer to my chest. Can't give away all my secrets. :happyno:

FWIW, my company installs mostly micros, a few optimizers, and a very few string inverters with add-on rapid shutdown boxes on residences in juridictions that enforce 690.12, but all string inverters on commercial buildings. It will be interesting to see what happens with commercial systems when the 2017 NEC becomes ubiquitously adopted, for some definitions of "interesting".

 

[SolarPro]

Remember that the requirements for controlling conductors within array [690.12(B)(2)] have delayed enforcement until January 2019. So the main difference between NEC 2014 and NEC 2017 for early adopting AHJs is that the 10' boundary outside the array shrinks to 1'. That certainly favors MLPE, but does not require MLPE.