Inverter sizing -- 2 x 6kW on 16.32 kW solar?

[vvuk]

The fun never stops with Tesla Energy. We finally got everything installed and got PTO, and I noticed that our production always caps out at a suspiciously exact 12.5 kW. I expected Tesla to install two 7.6 kW inverters (for a 7% undersize) or at least a 6+7.6. Instead we have two 6 kW SolarEdge inverters, for an inverter undersize of around 30%! I wish I had done a lot more research up front, but did not expect to need to become an expert on solar systems before ordering. (2 x 6kW was always what the system was designed with, so it's not a matter of them not delivering what was promised)

So -- is this actually a problem? I'm getting set up with a SolarEdge account so that I can at least see how much clipping we're getting. We have two sets of panels in slightly different orientations, so this could be a non-issue and only causing an actual loss during the hour or so that the sun is hitting both sets of panels at a good angle. Or each could be clipping at different times, and we're actually losing a good amount of daily production. I guess I'll find out soon. For those of you with 16 kW installs, what inverters do you have?

What's especially annoying is that the price difference between the 6 kW and 7.6 kW SolarEdge inverters is only around $175! Had I known what to ask for I would have pushed for the bigger inverters.

I should add: on sunny days (basically every week for the past week) -- we generally have about 100 kW total production daily. So we're still in good shape, though add in car charging and air conditioners and that gets eaten into!

 

[h2ofun]

The fun never stops with Tesla Energy. We finally got everything installed and got PTO, and I noticed that our production always caps out at a suspiciously exact 12.5 kW. I expected Tesla to install two 7.6 kW inverters (for a 7% undersize) or at least a 6+7.6. Instead we have two 6 kW SolarEdge inverters, for an inverter undersize of around 30%! I wish I had done a lot more research up front, but did not expect to need to become an expert on solar systems before ordering. (2 x 6kW was always what the system was designed with, so it's not a matter of them not delivering what was promised)

So -- is this actually a problem? I'm getting set up with a SolarEdge account so that I can at least see how much clipping we're getting. We have two sets of panels in slightly different orientations, so this could be a non-issue and only causing an actual loss during the hour or so that the sun is hitting both sets of panels at a good angle. Or each could be clipping at different times, and we're actually losing a good amount of daily production. I guess I'll find out soon. For those of you with 16 kW installs, what inverters do you have?

What's especially annoying is that the price difference between the 6 kW and 7.6 kW SolarEdge inverters is only around $175! Had I known what to ask for I would have pushed for the bigger inverters.

I should add: on sunny days (basically every week for the past week) -- we generally have about 100 kW total production daily. So we're still in good shape, though add in car charging and air conditioners and that gets eaten into!

For how long? 99% of the time, it will never clip. You will be fine. BUT, yep, folks looking at the details is what I have done. I have learned a lot. I made a mistake here and there, but, at least I put in 11.4K SE inverters so when I add more solar, the inverters do not have to change!!

 

[wwhitney]

So -- is this actually a problem?

It's only a problem if (a) you maxed out your available roof space and (b) you're not price sensitive and would be happy to pay more to produce more energy, even beyond the point of diseconomies of scale (*) (i.e. the marginal cost of additional kWhs starts going up.)

If both of those are true, then you could pay more for bigger inverters (and the cost would likely be a fair amount more than 2 x $175, as there are additional costs, like larger wires, possibly larger conduit, possibly larger panelboards, possibly more difficulty complying with the NEC panel backfeed limitations, possibly utility interconnection limitations, etc.)

Otherwise, even at your current DC/AC ratio of 1.36, it's likely that if you want more energy, you'd have a lower marginal cost/kWh to add more panels than to add more inverters.

Cheers, Wayne

(*) Edit: okay, that's not really diseconomies of scale, it's just hitting the roof size limit so that you are forced to have an economically suboptimal (closer to 1.0) DC/AC ratio.

 

[arnolddeleon]

I don't know how long you've been operating but also remember you will get dirt on the panels (you won't be cleaning it every day).

@wwhitney explains it very nicely. Mo inverters is not necessarily mo better.

This is a poor comparison but as data point I have over 17kW (DC rating) of panels on the roof various vintage and vendors and eyeballing the last few weeks I almost never exceed 11 kW at peak, a couple of days past 11 but under 12 kW. I'm not inverter limited.

@nwdiver would almost certainly say on a none roof space limited install you would have been better off spending the $350 on an additional panel.

 

[vvuk]

Ugh. Finally got proper access. The clipping is actually even worse than I expected. It's not even 11am and here's the input DC power to our two inverters -- and each one able to put out a max of 6kW AC. Actual AC production is currently 11.5 kW, when right now with 7.6kW inverters it would be at 15kW and would likely stay that way for the next 4-5 hours. This seems just a wasteful loss; probably around 25kW/day for 6 months.

 

[wwhitney]

I don't understand what your screenshots are supposed to be showing.

Does each inverter have all its panels on one plane? And can you generate a kW produced vs time of day graph for each inverter?

Cheers, Wayne

 

[holeydonut]

If your roof is maxed out, I can see why this is frustrating since you can't just add more panels to make up for clipping losses like some people can when their homes have huge roofs.

Can you show your production graph over a full day?

If it looks like this... I can see the argument that maybe you should have a larger inverter for the panels facing due south so you capture more.

But if it looks like this, the clipping isn't so bad, and dirty panels plus general decay is going to make this bit of clipping negligible.

As others noted, any larger inverter comes at a cost, and you may not recoup the value of the inverter over the life of your system. But since you're in PG&E territory with energy (for non CARES people) on track to be $0.45 per kWh next year... I feel like any extra solar harvest will be $ in your pocket in the long run.

 

[holeydonut]

I don't know how long you've been operating but also remember you will get dirt on the panels (you won't be cleaning it every day).

@wwhitney explains it very nicely. Mo inverters is not necessarily mo better.

This is a poor comparison but as data point I have over 17kW (DC rating) of panels on the roof various vintage and vendors and eyeballing the last few weeks I almost never exceed 11 kW at peak, a couple of days past 11 but under 12 kW. I'm not inverter limited.

@nwdiver would almost certainly say on a none roof space limited install you would have been better off spending the $350 on an additional panel.

You lucky bastards with having space to add more panels...

 

[vvuk]

Gah, I completely misread that diagram. Will repost updated numbers once I have a better graph.

 

[nwdiver]

This seems just a wasteful loss; probably around 25kW/day for 6 months.

Unlikely, for your location. 12kW of inverters vs 15.2kW. Oversize of 30% vs 7%. PV Watts estimates energy loss of ~20kWh/yr due to inverter saturation. ~20kWh per year... not per day. Even at the low cost of $175 to upgrade your inverter that's a ROI of over 40 years....

 

[h2ofun]

You lucky bastards with having space to add more panels...

IN 2 days I might have 88

 

[vvuk]

Actually pulling the proper data for today for the two inverters, graph attached. The right side of the graph doesn't have full data (it's still sunny , hence the sharp falloff.

Each inverter is clipping for 3.5 hrs on sunny days. But by how much is going to be hard to tell -- I can probably estimate a theoretical max based on the shape of the curve, assuming that (ideal, unlikely) I would get the full 8kw from each string at a peak. So rough back of napkin, assuming simple triangle at the peak.. 1.75*2*2, giving me 7 kWh/day. So, okay, 7%-ish loss on fully sunny days (we generate ~100 kWh/day right now). So, yeah, I guess I won't worry about it, but would have been nice if Tesla had sized this just a bit higher!

We do have more roof space, so adding a new string would be possible. We'll wait and see what 2 EVs in actual post-lockdown conditions do.

 

[nwdiver]

Actually pulling the proper data for today for the two inverters, graph attached. The right side of the graph doesn't have full data (it's still sunny , hence the sharp falloff.

View attachment 671917
Each inverter is clipping for 3.5 hrs on sunny days. But by how much is going to be hard to tell -- I can probably estimate a theoretical max based on the shape of the curve, assuming that (ideal, unlikely) I would get the full 8kw from each string at a peak. So rough back of napkin, assuming simple triangle at the peak.. 1.75*2*2, giving me 7 kWh/day. So, okay, 7%-ish loss on fully sunny days (we generate ~100 kWh/day right now). So, yeah, I guess I won't worry about it, but would have been nice if Tesla had sized this just a bit higher!

We do have more roof space, so adding a new string would be possible. We'll wait and see what 2 EVs in actual post-lockdown conditions do.

Looks like you were unusually cool in Sacramento today? That clipping will probably go away when temps go over 95F.

 

[h2ofun]

Looks like you were unusually cool in Sacramento today? That clipping will probably go away when temps go over 95F.

Yep, very cool! Going to be over 100 next week

 

[holeydonut]

Yep, very cool! Going to be over 100 next week

If you can get your solar running by then you’ll have enough energy to set your AC to 72F…