10Kw Grid Tie DIY

[DThames]

I put this system in June 2019, my first. We have great grid tie laws in Arkansas. Full credit for any excess that goes onto the grid. I get $$ credit on my account monthly with monthly energy cost adjustment. This credit pays for my monthly service access fee and a bit more. I oversized, planning to retire in a few years and expect to be using more power once I am at home all of the time. The mount is fixed angle. The substructure, I made from steel drill pipe and then used Iron Ridge mounting rails and hardware. I have 8, 4 port APSystems QS1 microinverters, peak total rating 9.6Kw. From the house is a 240v 50 amp circuit coming from my main panel, with outside disconnect. At the array field I have a post with breaker box and local 120v and 240v outlets. All normal 240v AC wiring ran right to the grid tie inverters without any "solar" gear in the system until I connect 240v to the microinverters. I have 32 older designed (but never used) SunPower P17-340 panels. On a cool sunny day I can hit 10Kw at midday. I am making about 1500 KWhr per month average.



Sorry little tree, but you will have to go.


Local 240v circuit runs to the inverters under construction.

 

[Deleted member 783]

WOW. First system??? You're good.

 

[Will Prowse]

Holy cow!! Very nice!

 

[erik.calco]

I put this system in June 2019, my first. We have great grid tie laws in Arkansas. Full create for any excess that goes onto the grid. I get $$ credit on my account monthly with monthly energy cost adjustment. This credit pays for my monthly service access fee and a bit more. I oversized, planning to retire in a few years and expect to be using more power once I am at home all of the time. The mount is fixed angle. The substructure, I made from steel drill pipe and then used Iron Ridge mounting rails and hardware. I have 8, 4 port APSystems QS1 microinverters, peak total rating 9.6Kw. From the house is a 240v 50 amp circuit coming from my main panel, with outside disconnect. At the array field I have a post with breaker box and local 120v and 240v outlets. All normal 240v AC wiring ran right to the grid tie inverters without any "solar" gear in the system until I connect 240v to the microinverters. I have 32 older designed (but never used) SunPower P17-340 panels. On a cool sunny day I can hit 10Kw at midday. I am making about 1500 KWhr per month average.

View attachment 2377

Sorry little tree, but you will have to go.
View attachment 2378

Local 240v circuit runs to the inverters under construction.
View attachment 2379

Very nice! Can you share a list of your mounting hardware, ideally with links if you have online? How deed were your holes?

I love the simplicity of your mounts. If that can hold winter, I'm in. How are your winds?

 

[Bob142]

Great job. You've set a pretty high bar for a first system install!

 

[DThames]

Very nice! Can you share a list of your mounting hardware, ideally with links if you have online? How deep were your holes?

I love the simplicity of your mounts. If that can hold winter, I'm in. How are your winds?

I will step back to the beginning to help this make good sense. I first studied tilt information to see what angle I needed. I looked at the previous 2 years power bills to see what my typical power consumption was. I then ran the PV Watts Calculator to see what size system I needed to get the desired annual output. I then went to the IronRidge Ground Based Design Assistant and configured the mount. The IronRidge design accounts for the typical snow and wind load of your area. You can override the values if desired. The resulting plans include drawings and Bill Of Materials ()BOM. The drawings include hole size and concrete mass/volume for the holes. The BOM includes retail pricing (scary), but you can send the BOM to IronRidge vendors for a real quote. I noticed the post top caps were fairly expensive from my perspective. Also the specified galvanized schedule 40 steel pipe was fairly expensive. In my area there is a lot of salvaged drill stem pipe. I found one size pipe would slide right over the 2-3/8" OD pipe that the rail to pipe mounting clamps were for. So my cross pipes are the same OD as specified, not galvanized, but with a thicker wall. The vertical pipes (made with Tees on top) are 2-7/8" OD, a bit larger than specified. My hole depth was about 42 inches and 12"-15" in diameter. I do have some wind protection from the north with the hilltop and some trees behind the panels. The design tool was figuring for 100 MPH winds. I tried to keep the basic design pretty close to what the IronRidge plans spelled out, so it would still have the strength as expected.

TeePipes under construction


Setting one of the lower cross pipes and supports. Dangle the pipes, pouring the concrete around them, and they will be lined up fairly well.

 

[erik.calco]

I will step back to the beginning to help this make good sense. I first studied tilt information to see what angle I needed. I looked at the previous 2 years power bills to see what my typical power consumption was. I then ran the PV Watts Calculator to see what size system I needed to get the desired annual output. I then went to the IronRidge Ground Based Design Assistant and configured the mount. The IronRidge design accounts for the typical snow and wind load of your area. You can override the values if desired. The resulting plans include drawings and Bill Of Materials ()BOM. The drawings include hole size and concrete mass/volume for the holes. The BOM includes retail pricing (scary), but you can send the BOM to IronRidge vendors for a real quote. I noticed the post top caps were fairly expensive from my perspective. Also the specified galvanized schedule 40 steel pipe was fairly expensive. In my area there is a lot of salvaged drill stem pipe. I found one size pipe would slide right over the 2-3/8" OD pipe that the rail to pipe mounting clamps were for. So my cross pipes are the same OD as specified, not galvanized, but with a thicker wall. The vertical pipes (made with Tees on top) are 2-7/8" OD, a bit larger than specified. My hole depth was about 42 inches and 12"-15" in diameter. I do have some wind protection from the north with the hilltop and some trees behind the panels. The design tool was figuring for 100 MPH winds. I tried to keep the basic design pretty close to what the IronRidge plans spelled out, so it would still have the strength as expected.

TeePipes under construction
View attachment 2394

Setting one of the lower cross pipes and supports. Dangle the pipes, pouring the concrete around them, and they will be lined up fairly well.
View attachment 2395

Thank you so much, @DThames !!! I'll let you know if I have questions later, which I almost definitely will. But, you are providing all the info I hoped for! It is going to be very helpful.

I can tell you as a skydiver that a bunch of trees next to an open area will create turbulence when the wind is coming from the direction of the trees. I don't know how this impacts load on solar panels. Perhaps it doesn't matter. I can only tell you that it's a big problem when landing with a canopy. Many won't jump when the turbulence is bad. For our drop zone, we preferred winds from the south which came across an open field that often had corn or soybeans. But, those winds can potentially be strong. Wind = brakes for us. But too much wind is bad, too. Turbulence can land you on your ass fast as it messes up the lift on the canopy and can suddenly turn in a different direction. Overall, though, from a solar panel perspective, the trees could be a net plus or negligible, because a panel is not trying to land. lol

Turbulence from trees isn't something we'd normally think about or notice if not trying to land. So just helping you to be aware of it.

This can also happen if wind goes over other structures like buildings. But, the wind seemed to be the choppiest from the trees. Buildings tend to at least have smooth uniform surfaces in comparison.

 

[DThames]

Thank you so much, @DThames !!! I'll let you know if I have questions later, which I almost definitely will. But, you are providing all the info I hoped for! It is going to be very helpful.

I can tell you as a skydiver that a bunch of trees next to an open area will create turbulence when the wind is coming from the direction of the trees. I don't know how this impacts load on solar panels. Perhaps it doesn't matter. I can only tell you that it's a big problem when landing with a canopy. Many won't jump when the turbulence is bad. For our drop zone, we preferred winds from the south which came across an open field that often had corn or soybeans. But, those winds can potentially be strong. Wind = brakes for us. But too much wind is bad, too. Turbulence can land you on your ass fast as it messes up the lift on the canopy and can suddenly turn in a different direction. Overall, though, from a solar panel perspective, the trees could be a net plus or negligible, because a panel is not trying to land. lol

Turbulence from trees isn't something we'd normally think about or notice if not trying to land. So just helping you to be aware of it.

This can also happen if wind goes over other structures like buildings. But, the wind seemed to be the choppiest from the trees. Buildings tend to at least have smooth uniform surfaces in comparison.

HaHa. I jump too. I jump mostly at Dallas and until last year I would make 2 trips per year to zhills. Things have changed the past year and I am closer to home more than I was.

 

[erik.calco]

HaHa. I jump too. I jump mostly at Dallas and until last year I would make 2 trips per year to zhills. Things have changed the past year and I am closer to home more than I was.

Awesome!! So good to meet another jumper here. lol

That means you have already thought about the turbulence. Have you concluded if it impacts the panels? I'm wondering how to consider wind in my backyard, and really am not sure. I only know that some of my lawn chairs get knocked over every winter.

 

[nifty-stuff.com]

This is fantastic! Thanks @DThames for the great details and pics!

I really wish I had oversized my first installation. Now I'm in the market for a 2nd system to augment what I already have.

 

[DThames]

One year update. Extra power is $$ credit on my bill to help pay the $22 per month meter service fee. My low credit over the winter was $4, so I just was able to not pay any out of pocket to the power company. I am closing the first year out with $95 credit on my account after paying my $22 this month.

After we did our taxes I determined I had a little over $13,000 in the system and got $3800 of that back in federal tax credit. So now it seems our R.O.I. will be just over 6 years.....and no power bill when I retire in a few years.

 

[analogue]

Did you get the ECU-R for the QS1 microinverters from APSystems? I'm curious about the web interface that runs on the ECU-R directly (not the cloud web interface) and if it exposes per-panel metrics in realtime. I'm trying to spec out my first system and I'd like to be able to scrape per-panel metrics with no reliance on the cloud/internet.

 

[DThames]

Did you get the ECU-R for the QS1 microinverters from APSystems? I'm curious about the web interface that runs on the ECU-R directly (not the cloud web interface) and if it exposes per-panel metrics in realtime. I'm trying to spec out my first system and I'd like to be able to scrape per-panel metrics with no reliance on the cloud/internet.

Yes I have the ECU-R. It has a button on it that when pressed will turn on a wireless radio for one hour. I set my phone wireless to that network and the phone app can access the ECU-R data. You can look at the data per panel but it would be difficult to collect or download it, based on what I have seen.

APSystems will not allow you to set up the ECU-R account. Without that account the ECU-R does not have the list of inverter radio IDs to collect data from. I purchased mine from Renvu and they helped get the ECU-R account set up so they would talk to each other.

 

[analogue]

Let me see if I'm understanding correctly. Once Renvu helped get the ECU-R account set up, you're able to connect from your phone to the ECU-R directly (ECU-R is the WIFI AP and your phone is the WIFI client) and get per panel stats on your phone with a little manual intervention (have to press the button on the ECU-R).

So I'm thinking if you're able to do that, then the other way to talk to the ECU-R directly without button pressing in a more automated way that is conducive to scraping is via the RJ45 port with a CAT5 cable hooked up to your home network. The ECU-R would get an IP address handed out from your home network's router/DHCP server and you should be able to point a web browser to the ECU-R's IP address and use it's web interface.

This should even work when your ECU-R is connected to your home's WIFI Router because that is how it is uploading data to APsystems cloud servers.

Or maybe I'm just missing something - did they remove the web server that runs on the ECU-R completely?

 

[DThames]

Let me see if I'm understanding correctly. Once Renvu helped get the ECU-R account set up, you're able to connect from your phone to the ECU-R directly (ECU-R is the WIFI AP and your phone is the WIFI client) and get per panel stats on your phone with a little manual intervention (have to press the button on the ECU-R).

So I'm thinking if you're able to do that, then the other way to talk to the ECU-R directly without button pressing in a more automated way that is conducive to scraping is via the RJ45 port with a CAT5 cable hooked up to your home network. The ECU-R would get an IP address handed out from your home network's router/DHCP server and you should be able to point a web browser to the ECU-R's IP address and use it's web interface.

This should even work when your ECU-R is connected to your home's WIFI Router because that is how it is uploading data to APsystems cloud servers.

Or maybe I'm just missing something - did they remove the web server that runs on the ECU-R completely?

I just now looked at my router, got the IP for the ECU-R device, and put the IP in the browser.....nothing opens.

The ECU-R setup involved, (weeks after receiving hardware)
1. send inverter address codes to Renvu
2. Renvu set up account and recorded inverter serial nbr/addresses assigned to ECU-R serial number
3. After my ECU-R was connected to the cloud account, it learned what inverters to listen for. Otherwise it ignores them.

The cloud account is free and works properly most of the time. Per panel data is updated every 5 minutes. All data, every 5 minutes. That data is available on the browser or with a different phone app.

 

[analogue]

Thanks for the info. This is an unsettling trend in consumer devices these days - everything is dependent on the cloud and customers are locked out of local access to their data. Of course, the same customers get left out in the cold when the company folds any number of years from now and the cloud interface disappears. Wonder if anyone has done an nmap scan on the ECU-R or poked at it to see if it exposes data on non-standard TCP ports.

Strike one for APSystems

 

[patdown45]

I am in the process of building my own 7.2kw grid tied system using TRINA 180w Da01 modules I have 46 and have started a small array using 8 in pairs of 2 in parallel to each of the 4 channels on my QS1 and have seen quite large disparities between the 4 channels even though all the modules are in direct sun without shading. channel 2 is always showing 2 watts which means no output, but when I check at the back of the panel as I can pull off the junction box cover and check the voltage and the amperage output at each pair as they are wired they are all putting out between 9 and 10 amps. is this overloading the QS1? Also I had one pair tied in using a set of short cables designed to parallel 2 into one and found that it was not allowing the amperage of both modules to be delivered? I replaced the cable with the solid one piece two to one connector and now all pairs are putting out the correct amount of power...but the ECU app is showing channel 2 is not and the other 3 are miss matched in the amount of power output? I am glad I did this test before I purchase another 3 microinverters as I believe there is some issues with these QS1 Micro inverters but now I have over a grand tied into these 3 and the trunk cable

 

[DThames]

Is channel 2 always not producing, even if you swap cables and swap panels? Can you make channel 2 work?

 

[DThames]

Also a clip on AC amp meter can tell you the real story. The ECU queries the inverters and records the data. You should be able to determine if it is a producing problem or a reporting problem by looking at the AC amps.

 

[Hedges]

One thing I see lacking in the structure is triangles. Yours depends on torsional rigidity/strength of the upright pipes which may be sufficient.
With 16 panels it is probably 600 pounds, not a huge weight.

My first installation was three, 2x4 arrays of 25 pound 120W Astropower on Unirack pole mounts.

https://www.altestore.com/store/i/multimedia/images/UTPM_image.png/maxx345/y250/dpw-universal-top-of-pole-mount-6-modules-60-cell-from-altEstore.com.png

As designed that didn't use diagonal braces as it had 6" pipe in a footing, but mine was bolted not set in concrete so I added braces.

Later replaced with ULA series large ground mounts. That uses all 2" pipe.
The ULA design has some diagonal struts going from ground to top of uprights in both X and Y direction. That provides greater resistance against wind and seismic loading.

https://www.websolarsupplies.com/wp-content/uploads/Unirac-ULA.jpg

Later additions I cobbled together use U-channel or other pipe clamps for diagonals.