Welcome to Tesla Motors Club
Discuss Tesla's Model S, Model 3, Model X, Model Y, Cybertruck, Roadster and More.
Register

Feedback on System Design - 17.6kW and 2 PW

This site may earn commission on affiliate links.
Hi everyone,

I wanted to get feedback on the layout of our system design, specifically how the solar strings are tied into the inverters. There are 6 separate arrays due to the rooflines and other obstructions that has panels facing all 4 directions.

My first observation was that the groupings as tied into the inverters made sense and it looked like they were utilizing as many inputs as possible of the inverters (2x4 available). After more reading and a closer look, it appears there are 3 separate groups that have jumpers (ex MP4 and MP2) vs being separately handled (MP5 and MP6). Any feedback on this specific setup? My hope would be the strings were as independent as possible since some face different directions (MP4 and MP2) or will have some shading effects (MP1).

Another related but different question is does the diagram provide any insight into how the arrays will be tied together coming from the roof? My original assumption was that each MP would have a separate ‘run’ to get to the inverter. This would mean no MPs would be in parallel or series prior to the inverter. I’m not sure that assumption is correct based off of other discussions I’ve seen despite how the diagram shows 8 separate ‘blocks’ coming into the inverter setup. For example MP1 is shown as 2x6 at the inverter but appears to have jumpers. Does that mean MP1 will have 2 sets of wiring running from the array (6 panels each) but then tied together at the inverter?

I appreciate any feedback anyone has and let me know if there are any additions l details needed. I did my best with the photos but let me know if there’s a better way to post them.

Thank you
2AC3018D-0F3D-4349-9443-0590411D5EF8.jpeg
A7820041-2E7E-44D1-A9E0-CABB78D33028.jpeg
 

Attachments

  • DBE57B44-3F3E-4C99-AF74-98D80788759C.jpeg
    DBE57B44-3F3E-4C99-AF74-98D80788759C.jpeg
    378.9 KB · Views: 121
  • A21D33C9-3294-4A84-B1F9-59F9F4663009.jpeg
    A21D33C9-3294-4A84-B1F9-59F9F4663009.jpeg
    419.3 KB · Views: 44
The largest two groups (2x6 and 2x9) are combined on the roof and will each have just 1 pair of wires running to the inverter.

In total there would be 5 pairs of wires coming down to the inverters.

I would ask MP2 and MP4 be separated rather than combined, making it 6 pairs of wires coming down to the inverters.
 
  • Like
Reactions: power.saver
14.4 kW system, no Powerwalls. Just trying to get a little power to offset the power used by our MX and MY cars, our house consumes much more power than this system will generate.

I also have to wonder why some of the panel groups are separately wired and some are tied together at the roof mounted string junction box, as there are a total of 8 inputs available on the two inverters.

The configuration has four groups of 6 panels each, of which MP4 and MP10 are paralleled at the roof junction box and paralleled again at the inverter's DC inputs, and of which MP7 and MP8 are also paralleled and run to the other inverter.

There must be a reason other than trying to save a few feet of wire from the junction box to the inverters?

The other groups, two with 4 panels and one with 5 panels are separately run to the inverter's DC inputs.

Complicated roof. Designed so as to not see panels from front of the house.


Screen Shot 2022-11-14 at 4.14.18 PM.png
Screen Shot 2022-11-14 at 4.10.20 PM.png
 
14.4 kW system, no Powerwalls. Just trying to get a little power to offset the power used by our MX and MY cars, our house consumes much more power than this system will generate.

Complicated roof. Designed so as to not see panels from front of the house.
What happened to the trend of wanting everyone to know that you had solar? Assuming that north is the top of the drawing nearly all of the panels are in sub-optimal to very sub-optimal locations which means that for a 14.4kW/36 panel systems you will be getting the "little power" that you desire. The six panels that you have on far right look like they will be in a shadow from the adjacent roofs and will be useless for most of the day.

The only saving grace to this layout is that you are in Boca Raton, Florida so even a due north system will get you 1.2 multiplier and your 14.4kw system should have at least a 17,000 kWh annual production. Personally, I wouldn't go forward with the layout as it is sub-optimal from what you could be getting by moving more of the panels to south facing roofs at no additional cost to you, but with significant increase in production which increases your savings.
 
Apparently not in our neighborhood. AFAIK I will be the first solar installation.

Tesla says 20,000 kWh annual production. Any thoughts as to my question as to why the four 1x6 groups are combined into two home-runs to the inverters?
Combining the groups together is usually done for cost and to simplify the installation. I would push for these to be independent runs to the inverter as the panel orientations are different for MP4/MP10 and MP7/MP8 are far apart on the roof already and I think MP8 will have a lot more shading than MP7 does.

20,000 kWh annual production is higher than I would have expected.
 
  • Like
Reactions: Eric33432
Tesla says no change can be made on the stringing of the panels into the inverters, that it is configured for optimal production. Perhaps when they actually install it they will change it, but I suspect the installers have to install per the drawings.

But a couple of errors were made on the electrical design. For example, my electrical service GEC is connected at the meter box, not at the main service panels, but Tesla's plan shows the GEC connected at the main service panels.

IMG_0640 2.JPG



The GEC is connected to the same set of lugs where the neutral conductors come in from the POCO (cables with yellow stripe), and firmly bolted (and bonded) to the cabinet. And there are only 3 wires leaving the meter to each of my three 200 amp main service panel disconnects, where the neutral is also bonded to the main service panel. From each main service panel there are 4 wires going to the load centers in the garage.

IMG_1873.JPG


I got into a "discussion" with an electrician I had at the house installing a new sub-panel to feed my pool equipment about this, and he wanted to run green wires from GEC at the meter to each main service panel, and remove the bonding jumper at these panels. It was like he was treating the meter as the main service panel, and the main service panels as sub-panels!

So I sent him the below video made by Mike Holt which says that it is correct. He admitted he had looked at it wrong and said the fact that I have three main service panels had temporarily confused him.

And I have checked several of my friends and family's houses in Boca and some have the GEC connected at the meter (with just three wires leaving the meter to the main disconnect), and some have the GEC going to the main disconnect. So apparently the City of Boca Raton electrical inspectors agree with Mike Holt.


Another error on Tesla's plans shows connecting the neutral to leg B in the Square D load center, and there are a couple of discrepancies in the wire that is specified, such as no neutral in the Cutler Hammer disconnect and Square D load center (which neutrals are shown on the plan).

So they cannot wire everything exactly according to the plan.

I don't plan on bringing any of this to Tesla's attention. I am sure the guys who do the installation will figure everything out. But if they want to run a green wire to my main service panels I will show them Mike Holt's video.
 
Last edited:
I don't plan on bringing any of this to Tesla's attention. I am sure the guys who do the installation will figure everything out. But if they want to run a green wire to my main service panels I will show them Mike Holt's video.


I just now noticed that you are not the original OP of this thread. In any case, I think you are making a big mistake with " I dont plan on bringing any of this to Tesla's attention", that is, if you expect it to be done differently when they are on site. I highly doubt any onsite installation crew from tesla is going to stop and look at what they would consider to be "some youtube video" (even if it is by Mike Holt) telling them to do something different than whatever they are doing, if they happen to be doing something that contradicts the video.

They also are not going to particularly want to hear a homeowner "telling" them "you are doing it wrong". All that is going to do is brand you as a difficult customer. You want to get all that ironed out before they show up, so they are installing as much as possible "to the plans they have".

Expecting something different is a recipe for your after installlation thread to be a "Tesla never listens to anyone!!!" rant.
 
@Eric33432 how sure are you sure that your neutrals aren't bonded through the bus bar behind the meter to the grounds? Sure looks like it.

All the best,

BG
The bar behind the meter with 8 lugs is a continuous bar which is bolted to the aluminum meter box

Screen Shot 2022-11-18 at 9.48.19 AM.png
The grounding electrode conductor (GEC), copper water system in the house, the two neutrals (yellow stripe) coming in from the POCO, the three neutrals going to the three main service panels, and on the far left a wire with some green tape on it that goes to bond the gutter (where the wires leave that leave through the top of the meter box and are run through the gutter to the main service panels) are all connected together at this point, which complies with the explanation of how to properly ground an electrical service in Mike Holt's video.
 
Last edited:
  • Like
Reactions: BGbreeder
I just now noticed that you are not the original OP of this thread. In any case, I think you are making a big mistake with " I dont plan on bringing any of this to Tesla's attention", that is, if you expect it to be done differently when they are on site. I highly doubt any onsite installation crew from tesla is going to stop and look at what they would consider to be "some youtube video" (even if it is by Mike Holt) telling them to do something different than whatever they are doing, if they happen to be doing something that contradicts the video.

They also are not going to particularly want to hear a homeowner "telling" them "you are doing it wrong". All that is going to do is brand you as a difficult customer. You want to get all that ironed out before they show up, so they are installing as much as possible "to the plans they have".

Expecting something different is a recipe for your after installlation thread to be a "Tesla never listens to anyone!!!" rant.

I have had similar experiences with professionals, such as the electrician I mentioned in my post above, and agree they do not want a homeowner telling them how to do their job. You make a very good point, and I will take your advise. Thank you.

But if they wire this up exactly the way the wiring diagram shows, and connect the neutral to the leg B bus of the new Square D panel they install, when they flip the new Cutler Hammer switch, the leg b fuse in the switch will blow, and they will know they have made a huge mistake. I do not expect them to do that, and that they will ignore this error in the plan.

I also note that the Cutler Hammer switch is wired as service equipment, not as a sub panel, so would expect the neutral to be bonded to the GEC assuming there is a neutral available.

Note the bill of materials does not even specify the neutral wire that is shown in the wiring diagram!

Screen Shot 2022-11-18 at 10.17.41 AM.png


So perhaps a neutral will not even be installed??? I note the inverters do not require a neutral. In that case, running a bare #6 copper wire directly to the ground rod would be fine because there is no neutral to bond.

Bottom line is there are inconsistencies in the wiring diagram/bill of materials that the installation crew will have to sort out, and in the end, the city electrical inspector will have to approve the job.
 
Last edited:
The bar behind the meter with 8 lugs is a continuous bar which is bolted to the aluminum meter box

View attachment 875814The grounding electrode conductor (GEC), copper water system in the house, the two neutrals (yellow stripe) coming in from the POCO, the three neutrals going to the three main service panels, and on the far left a wire with some green tape on it that goes to bond the gutter (where the wires leave that leave through the top of the meter box and are run through the gutter to the main service panels) are all connected together at this point, which complies with the explanation of how to properly ground an electrical service in Mike Holt's video.

My understanding is that the Service Entrance panel is the panel with the OCPD in it. Therefore that meter enclosure is just that and might as well be a hot box. I am not following super closely, but you might check out the diagram in section 230, which shows which enclosures are a service.

I have learned and seen that the main bonding jumper should only occur once inside each service panel, and not at the meter socket, but I am not 100% positive that there is not some provision for your installation to be code compliant, since I am not deep into your details.

@wwhitney might be able to provide some insight.

As far as the errors in your line diagram, I'd recommend pointing out the issues you see to the installers. I am 100% sure they will not wire up the neutral to leg B for very long. Tesla probably is banging these out pretty fast, so your errors are just that. It is a very tiny chance that the neutral would get missed (it is shown visually), most electricians know better and will size the neutral the same as the conductors.

You can see that who drew this did show bonding of the Neutral and Ground inside the AC Disconnect. I think that technically none of the PV needs a Neutral to work, but also that any subpanel should have a neutral, even if it is sized at the minimum which I believe is no smaller than the EGC.
 
  • Like
Reactions: Eric33432
on the far left a wire with some green tape on it that goes to bond the gutter (where the wires leave that leave through the top of the meter box and are run through the gutter to the main service panels)
I'm not clear if the separate wire to bond the gutter on the utility side of the service disconnects is actually correct. Equipment on the utility side of the service disconnect(s) is to be bonded to the neutral, so normally one or all of the 3 service neutrals would be bonded to the gutter. Sounds like the 3 neutrals pass through unspliced and this extra neutral/supply side bonding jumper is being used to bond the gutter. [On reflection, this is probably fine.]

With the PV installed you'll effectively have 4 different service disconnects. While the one for the PV is often called a PV disconnect for a line-side PV connection, it can be wired just like a service. The 2020 NEC has removed most of the ambiguity on how to handle the PV disconnect by specifying it needs to be wired just like a service (new section 250.25). Each service disconnect (including the PV) will have a main bonding jumper in it and originate an EGC system.

The GEC(s) can be landed at or upstream of the service disconnect(s). If you were forced to land your GEC at the service disconnects, then you'd need a common GEC with jumpers to each of the 4 service disconnects. That are you allowed by your power company to land it in the meter is a great simplification, as you just need the one neutral-GEC connection.

BTW, one side effect of having multiple services with multiple main bonding jumpers in the same building is that if/when the separate EGC systems become interconnected, you will end up with some neutral current on the EGCs. This is basically unavoidable given the rules. If that concerns you at all, you could mostly isolate the phenomenon to the vicinity of the 3 (now 4) service disconnects by intentionally tying all the EGC systems together there.

Where there other questions?

Cheers, Wayne
 
Last edited:
I'm not clear if the separate wire to bond the gutter on the utility side of the service disconnects is actually correct. Equipment on the utility side of the service disconnect(s) is to be bonded to the neutral, so normally one or all of the 3 service neutrals would be bonded to the gutter. Sounds like the 3 neutrals pass through unspliced and this extra neutral/supply side bonding jumper is being used to bond the gutter. [On reflection, this is probably fine.]

With the PV installed you'll effectively have 4 different service disconnects. While the one for the PV is often called a PV disconnect for a line-side PV connection, it can be wired just like a service. The 2020 NEC has removed most of the ambiguity on how to handle the PV disconnect by specifying it needs to be wired just like a service (new section 250.25). Each service disconnect (including the PV) will have a main bonding jumper in it and originate an EGC system.

The GEC(s) can be landed at or upstream of the service disconnect(s). If you were forced to land your GEC at the service disconnects, then you'd need a common GEC with jumpers to each of the 4 service disconnects. That are you allowed by your power company to land it in the meter is a great simplification, as you just need the one neutral-GEC connection.

BTW, one side effect of having multiple services with multiple main bonding jumpers in the same building is that if/when the separate EGC systems become interconnected, you will end up with some neutral current on the EGCs. This is basically unavoidable given the rules. If that concerns you at all, you could mostly isolate the phenomenon to the vicinity of the 3 (now 4) service disconnects by intentionally tying all the EGC systems together there.

Where there other questions?

Cheers, Wayne
Thanks, Wayne, for this.

Presumably what I have now was inspected when the house was built in 2006, (I purchased it in 2011) and nothing has been changed between the meter and the three service disconnects, which are in actuality Service Equipment rated Generac transfer switches, each with a 200 amp breaker for the utility connection. Tesla simplified the drawing by just showing them as "3x main service panel". The three neutrals leaving the meter (the wires with the white tape on them) simply go into a gutter mounted above the meter box and the transfer switches and are landed in the three transfer switches where they are bonded to the transfer switch cabinet and where the EGC is created. The wire with the green tape on the far left of the ground/neutral bar grounds the gutter in case of a fault, so in effect it is the same as if one or more of the neutrals was tapped for this purpose. Seems a cleaner way to do it, and if all three neutrals were tapped for this purpose then in effect they would be in parallel up to the tap point, which may not be desirable, I have no idea. BTW, everything is connected with PVC conduit.

Re the "PV disconnect", what you say makes sense. The question is: will Tesla install a neutral into this panel? According to the diagram, they will, but according to the bill of materials, they won't. Either way, it will probably be OK since the downstream panel is just for Tesla inverters which don't use a neutral. The problem would be if someone hooked a circuit to this panel that needed a neutral, but I expect there will be a warning table on the panel to not do that.

As you say, having the GEC landed at the meter is a great simplification since the local AHJ allows that. The question is will Tesla installation crew continue that or install per their wiring diagram. It will be interesting to see how they do that.

The EGCs at the load centers are definitely interconnected. There is a bit of so called "unintentional current" on the EGCs but as you say it is unavoidable, and the current on the GEC is just the difference in the ground resistance and the utility neutral conductor, ad is inherent due to the rules as you say.

The only question I have is: is the ground wire for the gutter large enough? It appears to be about 1/0 copper. Since all the wiring in the gutter is before any OCPD, should there be a fault, would it conduct enough current to cause the transformer's protection to trip, or would the wire just melt?

Can't think of any more questions right now.
 
Last edited:
I think that with the current, pun intended, NEC rules on grounding, ground loops are basically inevitable if you have more than one panel. I have seven panels, with six different grounding systems, per NEC. With the distances involved, and dry soils, I expect some amount of voltage on my grounds as a result. As a result, I favor double insulated equipment over grounded equipment when I can, and I avoid ground fault devices where permissible.

YMMV...

All the best,

BG
 
Re the "PV disconnect", what you say makes sense. The question is: will Tesla install a neutral into this panel?

Ah, I see that Florida is using the 2017 NEC, so that version provides some ambiguity on that. They must provide a fault clearing path for the EGC originating at the PV disconnect. So that either has to be a neutral service conductor that is bonded with a Main Bonding Jumper to the EGC, or grabbing an EGC from one of the other service disconnects that has a Main Bonding Jumper. I would suggesting asking them for the former, since that is cleaner IMO and is what 2020 NEC 250.25 requires.

As you say, having the GEC landed at the meter is a great simplification since the local AHJ allows that. The question is will Tesla installation crew continue that or install per their wiring diagram. It will be interesting to see how they do that.
Did not look closely enough at their wiring diagram, but nothing they are doing should require them to touch the GEC, given that you have everything landed at the meter. I suggest reminding them of this.

The EGCs at the load centers are definitely interconnected. There is a bit of so called "unintentional current" on the EGCs but as you say it is unavoidable, and the current on the GEC is just the difference in the ground resistance and the utility neutral conductor, ad is inherent due to the rules as you say.
Well, there are two different issues. The GEC can cause the service neutral to carry current due to an earth potential difference between the service transformer and the house grounding electrodes. That's definitely unavoidable.

But having multiple main bonding jumpers at multiple services with the resulting creation of multiple EGC system that then become interconnected somewhere downstream will cause load neutral current to travel on those EGCs. Since the intended neutral current path is in parallel with the path of MBJ to EGC to other EGC system to other MBJ back to neutral.

The only question I have is: is the ground wire for the gutter large enough? It appears to be about 1/0 copper.
See NEC 250.102(C). It goes by the size of the ungrounded service conductors. I'd go by the larger of (a) the two parallel sets landing on the grid side of the meter and (b) the three non-parallel sets on the load side of the meter. It's possible that because the sets on the load side of the meter aren't actually in parallel, you don't have to add all three up.


Cheers, Wayne
 
Just a follow up to all the above.

Tesla finally installed my solar system, in late June.

Before they would issue the building permit, my city required Tesla to modify their wiring plan, (the areas highlighted) to include the neutral wire between the meter and the new solar Service Equipment ("safety switch") and a neutral was run from the safety switch to the load center for the solar system. This was originally shown on the wiring plan but the neutral wires were not included in the bill of materials section of the drawing.

Regarding the GEC, the electrician grounded the safety switch exactly in accordance with the Tesla wiring plan: He ran a bare #6 GEC directly from the ground rod to the safety switch, and bonded the neutral and the GEC at the safety switch. This will create a parallel path for any neutral current through the GEC from that switch (to the ground rod, and then through the GEC from that same ground rod that connects to the meter).

The drawing shows my existing electrical service equipment grounded to the ground rod but that is wrong because the GEC for my three existing service disconnects connects at the meter, not the service disconnects. There are only three wires running from the meter to the service disconnects, which neutrals are bonded to the service disconnects, as was discussed in some of the posts above. But since there are only the Tesla inverters connected to the new load center, which do not use neutrals, it really does not matter.

The city did not catch the error where the B leg was connected to the neutral in the load center (or did not care), and of course the electrician did not wire it that way.

Everything else was wired exactly per the wiring plan.

The inspector did not care about any of this, and I did not say anything to him (or anyone) about my concerns. It all works very well, producing 70 to 75 kWh on clear days. Tesla left the system turned on and no one seemed concerned that we did not yet have permission to operate from the utility. My system has 37 400 watt panels and two 7.8 kW inverters so it will never clip. It peaks at a little over 12 kW.

I have noticed that the dashes on the meter, which normally move from left to right, are sometimes (but rarely) moving from right to left. If I turn on the generator, which disconnects my house from the grid, but leaves the solar system connected to the grid, the dashes move quickly from right to left when the system is producing 12 kW. Perhaps I already have a bi-directional meter (it is a Class 480 meter), but I have received one bill from FP&L and it did not show anything unusual except my bill was lower by about $350. Electricity costs 16.6¢ per kWh here, including the various taxes that are imposed in Florida.

Tesla's crew was very professional, they said my job, due to my tile roof, was the a difficult job. They started Monday morning and finished Saturday about 4 PM. Most houses in this part of South East Florida have tile roofs, the crew was based in Tampa and my job was the first they were doing in my area. They said they were moving on to another job in my area, I expect they will find their work difficult in this area. My system had an extra surcharge of $8,000 due to the tile roof type. But I am getting the 30% tax credit.



Screen Shot 2023-08-13 at 11.29.55 AM.png


Electrically the job was simple as there is no gateway, no batteries. I have a 60 kW generator and would need well over 100 panels to fully replace the amount of power that we consume. I may order some additional panels with the new 30% tax credit, which will be with us for several years as I understand. I think some additional inverters could be installed above the gutter which goes above the meter to the three Service Rated Generac transfer switches. The electrician had to move the irrigation controller which was in the way of installing one of the inverters. It is now right below the conduit to the right of the meter.
IMG_3385.JPG


IMG_3387.JPG

The three transfer switches feed three panels inside one of my garages. The little panel to the immediate left of the meter is for pool equipment and is fed from the load side of the transfer switch to which it is connected.

Any comments would be welcome.

Thanks for reading,

Eric
 
Given this:
the GEC for my three existing service disconnects connects at the meter, not the service disconnects.

Then this is unnecessary, as the service neutral conductor is already earthed:
Regarding the GEC, the electrician grounded the safety switch exactly in accordance with the Tesla wiring plan: He ran a bare #6 GEC directly from the ground rod to the safety switch

But this part is very important, as each service disconnect initiates its own EGC, and that EGC system needs to be bonded to the neutral to function properly:
bonded the neutral . . . at the safety switch.

Like any building with multiple services, the multiple EGC systems with multiple EGC-neutral bonds will allow neutral current on the EGCs if the different EGC systems ever become interconnected, as is not unlikely.

My system has 37 400 watt panels and two 7.8 kW inverters so it will never clip. It peaks at a little over 12 kW.

Given that you have 15.6 kW of inverter capacity, and currently hit at most a bit over 12 kW, you could increase the DC size by a 1/3 (add 12 more panels) without adding any inverters or having more than say 0.1% clipping losses for the year. And that's conservative for the case that all 37 or 49 panels are all facing the same direction; if you have multiple roof planes with multiple orientatoins, you could probably add more than that. [I did not review the thread, so perhaps info on the panel arrangement is already above.]

Cheers, Wayne
 
  • Like
Reactions: Eric33432