I've now had a chance to look at the supposed cost and LCOE advantage claims in the new Tesla inverter and - as I suspected last night - there is a whiff in the marketing speke.
It seems to me that the competitor inverter that Tesla most has in mind when doing its String+Optimiser calculations is the SolarEdge 8.0kW HD with the associated DCDC modules, aka "optimisers" that are mounted behind the solar PV panels themselves (which are also called modules).
SolarEdge??s latest generation of single phase inverters are designed using a novel power conversion technology that is based on a distributed switching and powerful DSP processing
www.cclcomponents.com
As it happens we are currently installing the 5kW version of this right now at my GF's property, to run approx 7kW of panels. Yes, 155% oversizing is allowed, and one would tend to do that on a shaded site or a multi-angle roof site to reduce LCOE as you'd never get all panels at max simultaneously. That's why we are doing it on my GF's. (I don't normally get directly involved in hands-on-tools work but I've been helping with some aspects of this job.)
But let's assume we run the mooted 8kW system cool by installing a pukka 8kW inverter. So my installer and I pulled up the price he would give me for the 8kW version and it is GBP 1300. These are kosher rates, not mates rates. To that we must add in 20 optimisers required to run a 8kW system using 400W panels, so 20 x GBP 40 = GBP 800. Now I also added on GBP 80 for these vague "supporting equipment" that Tesla claim are required, though neither of us can think of anything beyond perhaps a GBP 30 wifi aerial. So we have GBP 50 in hand in case anyone is arguing. Total = GBP 1300 + 800 + 80 = GBP 2180.
GBP 2180 = USD 2647, so let's call that $2650 to again be conservative in Tesla's favor.
Now let's pull up that Tesla cost graphic in the Tesla marketing brochure. I've added a scale bar on the cost axis (green blocks) and plotted the line for $2650, and we can also see that Tesla are suggesting that this system should cost $2810. It turns out that is a 6% cost overstatement in Tesla's advantage. Funny that.
View attachment 894845
That of course means that the LCOE calculation will shift by 6%. (for now assume those brown optimiser failure bars are correct, we'll come back to that)
View attachment 894846
And that 6% in turn means the Tesla claimed advantage disappears on almost all sites.
View attachment 894849
Regarding optimiser failures my personal experience is that I've had
2 1 failures on 24 panels in 7-years running across 2 sites so far. (Actually I can definitely remember one - and I've now checked with the installer ... it was only the one). Replaced under warranty in a few weeks - the job was done in less than an hour whilst the installer was passing by between other jobs. That's an
8% 4% failure rate. Tesla seem to be broadcasting FUD about 50% and 100% failure rates but I can't quite see how they run the calc on that. You can make your own call.
Regarding ease of set up, test, and repair what could be better faster ad easier than a one-per-module optimiser that tells you precisely how each solar panel and each optimiser is operating at all times, and alerts you to issues in either the install or during operation. As opposed to suspecting a problem somewhere on a string .... now get on the roof and find it .......
Now I can fully understand why Tesla might want to get into the inverter manufacturing game, and that they might want to do it with a string inverter in this manner. And that the result might be lots of marketing horlicks being sprayed around by Tesla.
But as investors we should see through that.
(plus I still stand by everything in my earlier posts re the strategic and decision-making / very-senior-management implications)