Help designing system for Tigo Optimizers - over current?

[jacob.wesley.smith]

Hello,

I currently have 30 335W panels (Yingli YL335D-36b) that I picked up for $30 a piece on facebook marketplace. I'm installing them on my roof and concerned about being over amperage for Tigo Optimizers. I know I need RSD, and the optimizers seem to be ~$800 more for the whole system, and as a data geek, I like the prospect of knowing exactly what my system is doing.

I'm planning on running all 30 panels to a 6000xp inverter. Due to the topography of our roof, they will be on 3 different strings - I'm hoping that I can combine 2 of the 3 strings by way of using the optimizers. However, I think that may bring my voltage up too high.

6000xp max PV Voltage: 500V
Indianapolis adjusted lowest temperature Voc for 10 panels: 509 V (I'm potentially okay with this and just turning off PV in the event of super cold (once a decade) weather)
Isc: 9.8 amp

So, my previous plan (prior to looking at Tigo data sheet) was to combine 10 panels in series, then parallel that with another 10 to stay right around 500 V max, with ~19.6 amps. However, it looks like the Tigo optimizer has a max of 15 amp.

Am I correct that doing 10s2p for the panels would lead to too much amperage going through the later optimizers and causing them to fail / be unsafe?

The simplest solutions seems to be to run 4 strings, but then I would need either a second 6000xp or a separate charge controller that went straight to the battery that the inverter could pull from.

Any advice on reasonable steps? I have only purchased the solar panels and 6000xp so far, so open to other options for what to look at.

Overall goal of the system is to replace ~half of our energy consumption; more in summer, less in winter. (aka, not fully off grid, nor do I want to mess with making it able to backfeed the grid, hence the 6000xp as the inverter of choice to pull from grid if we use more than the system has available).

Thanks!

 

[Quattrohead]

30 335W panels

Too many for a single 6000XP no matter how they are setup.

combine 10 panels in series, then parallel that with another 10 to stay right around 500 V max, with ~19.6 amps. However, it looks like the Tigo optimizer has a max of 15 amp.

You would be combining the strings after the optimizers so that is OK. BUT 10 in series is too risky. 2x8 and 2x7 is the way to go.

 

[zanydroid]

I would recommend pasting the panel specs to help catch more math and conceptual errors.

Indianapolis adjusted lowest temperature Voc for 10 panels: 509 V (I'm potentially okay with this and just turning off PV in the event of super cold (once a decade) weather)

Terrible design. Remove one panel from each string and leave that as spares in case of damage.

Turning off PV also means remembering & opening at the disconnect. "Turning off" could include just turning off the inverter at a power switch without isolating it. Which would not achieve your goal.

So, my previous plan (prior to looking at Tigo data sheet) was to combine 10 panels in series, then parallel that with another 10 to stay right around 500 V max, with ~19.6 amps. However, it looks like the Tigo optimizer has a max of 15 amp.

Am I correct that doing 10s2p for the panels would lead to too much amperage going through the later optimizers and causing them to fail / be unsafe?

(Oops, quattrohead already had a much more parsimonious explanation)

I don't understand why this would mess up the optimizers. One string of 10s will be within the optimizer limit at every point in the string.

Two parallel strings of 10s will not send 2x current through an optimizer.

There may be call to look at the max fuse rating on the optimizer, but as long as you are using normal electrical range panels (which we can guesstimate is the case based on the 335W power rating, but as above, you didn't paste the panel specs), it is likely fine.

Are you doing something more exotic than this / does 10s2p mean something different to you? Drawing a picture helps to disambiguate when we don't know how much to trust your engineering chops (if I work with someone daily at the office I feel more comfortable seeing less work shown, because I've established a baseline for their knowledge level and detail-orientedness level).

Additional SCC isn't that expensive esp when potentially installing non-code compliant combination of equipment -- in a decent fraction of states 6000XP doesn't have a cost-effective code compliant setup.

(aka, not fully off grid, nor do I want to mess with making it able to backfeed the grid, hence the 6000xp as the inverter of choice to pull from grid if we use more than the system has available).

6000XP has to be able to serve the full load of the output, otherwise it bypasses to serve 100% of the load from grid. So you may need to plan on moving circuits, and installing energy meter instrumentation.

I'm of the camp that, if you're going through all the trouble of learning how to install a (mostly?) code compliant DC system and doing it, you should verify how easy/difficult it is to interconnect a grid tie system, and what the net metering terms are. Granted, I'm in a solar friendly state so I assume all states are friendly (which is not true).

 

[jacob.wesley.smith]

Thank you both for your comments. I left out some of the details of my plan for brevity, though that may have caused more confusion than time savings.

My plan is to have a subpanel fed by a transfer switch - either from grid or from the 6000xp. Then, on my main panel, running a double 50 amp breaker to the 6000xp, so it can switch to bypass mode if necessary. That way, I should have no possible path for energy to backfeed the grid itself - anything in the subpanel would be fed by solar and, if/when batteries/solar are not enough to supply the load, it would kick back over to the grid connection of the inverter and still power those loads.

My conceptual mistake on the 10s2p comment was thinking that those parallel connections would be made before one of the optimizers, not after it. That makes much more sense - though I do also appreciate the comments that I should steer away from 10 panels in series regardless. I know I would need to install a DC breaker and turn that off in extreme cold weather, but how could I expect others to (wife, kids, etc.), and there's no real reason not to just do it the right way first.

As far as I know, the whole plan is to code in Indiana - is there a part I may need to rethink? I'll plan on a 2x8 and 2x7 and either a charge controller or second 6000xp. The second 6000xp is probably overkill, since I would almost never be pulling more than 50 amps on those subpanel circuits (I've installed an energy monitor to get a better idea of my baseline to try and size a system properly).

Thanks again

 

[zanydroid]

My plan is to have a subpanel fed by a transfer switch - either from grid or from the 6000xp. Then, on my main panel, running a double 50 amp breaker to the 6000xp, so it can switch to bypass mode if necessary. That way, I should have no possible path for energy to backfeed the grid itself - anything in the subpanel would be fed by solar and, if/when batteries/solar are not enough to supply the load, it would kick back over to the grid connection of the inverter and still power those loads.

OK, that makes sense. The 6000XP is a good choice for this since I believe it is pretty strongly guaranteed to not backfeed up the AC input in any mode.

I know I would need to install a DC breaker and turn that off in extreme cold weather, but how could I expect others to (wife, kids, etc.), and there's no real reason not to just do it the right way first.

I think most people use IMO disconnects instead of DC breaker. They are simpler and most likely more robust.

IMO a major advantage of solar + batteries vs other kind of power generation is that it has very few moving parts & maintenance. Manual steps take away that niceness.

As far as I know, the whole plan is to code in Indiana - is there a part I may need to rethink? I'll plan on a 2x8 and 2x7 and either a charge controller or second 6000xp. The second 6000xp is probably overkill, since I would almost never be pulling more than 50 amps on those subpanel circuits (I've installed an energy monitor to get a better idea of my baseline to try and size a system properly).

You can consider overpaneling 6000XP's MPPTs if you mostly need the 30x330W to get enough in winter. The 6000XP spec sheet does say 10000W is "recommended maximum". 2x8 and 2x7 is fine for 6000XP and below 10000W. I think the 7s2p voltage is high enough to max out the 6000XP MPPT limit. Max 6000XP charging is 6kW from solar.

You could also start with 6000XP, overpaneled, prewiring the DC in a way that's conducive to reconfiguring it feed more MPPTs down the line.

Here is the ESS section for Indiana. It has a somewhat similar UL9540 requirement to my state (either UL9540 or recycled EV batteries; my state has removed the small battery carveout). 6000XP with PowerPro is not a UL9540 combination (UL9540 is for the most part a system level certification comprising a specific set of inverters and matching batteries).

R327 (if you want to look for explainer documents from other states, it's R328 in states with a newer version of this code)

CHAPTER 3 BUILDING PLANNING - 2020 INDIANA RESIDENTIAL CODE

ICC Digital Codes is the largest provider of model codes, custom codes and standards used worldwide to construct safe, sustainable, affordable and resilient structures.

codes.iccsafe.org

This is an interesting amendment that includes older language, along with an update to allow V2H/V2L.

I'm not familiar with which SCCs are fully code compliant (AFCI, GFP, UL9540). Maybe MidNite (which might cost half of a 6000XP). Victron doesn't have AFCI. SignatureSolar's HV one don't have a listing.