diy solar

diy solar

My first Solar project

jsmalone1

New Member
Joined
Aug 30, 2021
Messages
21
So after 45 years as an electrician I decided to install a solar/ battery system at my house. My goal is really to have a whole house generator, as so many friends installed traditional generators, either natural gas or propane fueled, and I feel that is not the way I wanted to go. While the risk of not being able to obtain fuel isn't a risk to most people, it is to me. My community does not yet allow solar panels on our roofs so I installed 10 440 watt panels on stands on the south side of my house. I installed 23,000 wh of lithium iron phosphate batteries and a 5,000 watt Growatt inverter. So far I'm powering my entire house less the AC, water heater, dryer and oven. Being an electrician helped immensely with wiring a sub panel, transformer and the inverter. Beyond that however, solar and lithium batteries are a new learning experience for me. I'm installing a heat pump water heater and plan on adding this to my solar loads with a second 5,000 watt Growatt inverter. With little experience , I'm learning that more solar panels are probably needed. Any advice is welcome. I have an Emporium load monitor so I can see all my individual loads. Reading tips here really helped and I appreciate it.
 

Attachments

  • IMG_20211029_161039.jpg
    IMG_20211029_161039.jpg
    268.4 KB · Views: 43
  • IMG_20211102_142012.jpg
    IMG_20211102_142012.jpg
    106.4 KB · Views: 44
  • IMG_20211109_131648__01.jpg
    IMG_20211109_131648__01.jpg
    548.1 KB · Views: 43
First, is that an XT500 on the wall? My dad has a '76, and I rode the crap outta that thing until about 2000.

I would be VERY concerned about shading of the lower half of those bifacial panels from the fence. Depending on the circuit path in those panels, you could be taking out the entire half if you shade just the lower row of cells.

I would also hope you've given great attention to how you paralleled those dissimilar banks and confirmed that you are getting suitable distribution of the load/charge proportional to their size via clamp DC ammeter. The two circuit breakers on the wall presumably for isolating/protecting each bank individually cast doubt.

Additionally, I do not see BMS leads on the prismatic bank. I see that you paralleled 9 of them. I pray you did a proper top balance of all 16 and plan to immediately add a BMS last week. If not, I hope you are checking your cells constantly during charge.

The amount of solar availability and installed PV determine how much energy you can use on an average daily basis. The battery capacity determines how long you can go without charging.

I'm guessing the inverter has a ~80W idle load, which is 2kWh/day or 8.3% of your battery capacity. If you add another 5kW inverter, you lose another 8% of your battery capacity.

Heat pump water heaters are definitely way better than standard, but they "finish" by a pure heating element. Even if you use less total energy overall, you will have a big hit as it's hitting peak temps. I assume that's why you're looking at another 5kW unit.

Lastly, HOA restrictions on solar installations have been successfully challenged and 25 states now prohibit HOA from preventing solar installations. I couldn't find anything about Arkansas, so you might want to check with these folks:


As an electrician, I would presume that you are familiar with code. I would expect you are subject to NEC2017, which requires that each roof mounted panel have it's own cut off. This requires the addition of an optimizer with cut-off function for each panel.
 
Thank you for your response. So the prismatic cells were not completed in that photo. I have all 16 in series now and am waiting for my BMS. So I'm not using that bank of batteries. Our HOA is changing the rules, it was the developer. I don't understand the voltage specs of my panels. Solarever, specs say max power volts (Vmp) 40.99v open circuit volts (Voc) 49.62v. So I initially had all 10 panels in series and I was seeing 450+ volts often, which is the limit of the Growatt. I disconnected 1 panel and guess I will wire 2 in parallel??
 
Looks like the fence would shade the panels some hours and seasons.
How about securing the lower end near the top of the fence?
And the upper end out far enough that the eaves won't shade it (maybe hang from the eaves)?

A priority switching scheme to avoid multiple heavy loads at once should let you go farther with less inverter. You might get to run those heavier loads.
SoC based control might avoid wasting available power when battery is full, and avoid draining it too far.
For motor loads, starting surge will be the limiter. Do you know or can you measure inrush? I assume 5x nameplate current, but also got a cheap meter with inrush recently (previously used oscilloscope and current transformer.)
 
Thank you for your response. So the prismatic cells were not completed in that photo. I have all 16 in series now and am waiting for my BMS. So I'm not using that bank of batteries.

Glad to hear it!!!!

Our HOA is changing the rules, it was the developer.

Glad to hear it!!!!

I don't understand the voltage specs of my panels. Solarever, specs say max power volts (Vmp) 40.99v open circuit volts (Voc) 49.62v. So I initially had all 10 panels in series and I was seeing 450+ volts often, which is the limit of the Growatt. I disconnected 1 panel and guess I will wire 2 in parallel??

You have to use Voc when calculating the max voltage the controller will see. Voltage also goes up about 0.3% for every 1°C below 25°C. On a 50°F morning with cool panels, you'll see a 4.5% voltage increase ABOVE Voc, so even 9S would be too much.

You'll need to wire two strings of 5S and then parallel those strings into the charge controller, 5S2P.
 
We live near several lakes, many dock owners are switching to solar. As I was educating myself I said to a friend there that the systems you are putting on the docks are just toys. The more I learn is that comment was not far off.
 
Glad to hear it!!!!



Glad to hear it!!!!



You have to use Voc when calculating the max voltage the controller will see. Voltage also goes up about 0.3% for every 1°C below 25°C. On a 50°F morning with cool panels, you'll see a 4.5% voltage increase ABOVE Voc, so even 9S would be too much.

You'll need to wire two strings of 5S and then parallel those strings into the charge controller, 5S2P.
Great info concerning the heat pump water heater. I have not read that yet.
 
Note from:


It lists a 4500W heating element.

A heat pump is hard pressed to push water to 130°F for water heaters, but it's very effective for getting it most of the way. If you can be patient, it can eventually get you there; however, ambient conditions (cold heat sink) and demand may exceed its ability to perform to customer expectations, so it has a backup element.
 
I really appreciate your expertise. Coincidentally that is the water heater I bought, 50 gallon. So the intake will average around 60 degrees winter, 80 summer. Exhaust will be vented out to attic. There are only 2 of us here the vast majority of the time. Guessing, I would say most of our electric bill is the water heater, especially after I installed the Emporia monitoring system, I can see that but never really kept data. My lack of knowledge at this point plays into a question of 1. How much power will this water heater actually use with the 2 of us?
And 2. Is it even worthwhile to connect to my solar system?
As my original premise is to have a home generator not requiring refuelling from outside sources do I add this to my system. You made the most pertinent point that adding a second inverter my idle current draw alone is significant. I'm thinking about installing a second 5kw Growatt in parallel for backup in case of equipment failure AND that would give me the ability to experiment. I want to get some running experience before I start adding. Tomorrow I will rewire my panels per your instructions, thank you again. I'm waiting for an Overkill brand 48volt BMS, backordered.
Another area I neglected was to compute the solar required to keep my batteries charged, my home load and how much battery I actually need. Essentially I bought what I "thought" would be right. Another issue you and another person pointed out is the shading of my panels. 10 weeks there is a shadow from the fence, average 20% of the physical panel. I humbly admit I don't know the consequences of this. The panels don't shut down, oddly, at least to me, the panels generate power as soon as the sky brightens up in the morning and do very well on cloudy days. I did not know this. If there is a method of measuring this performance and associated loss I would appreciate that direction.
 
I really appreciate your expertise. Coincidentally that is the water heater I bought, 50 gallon. So the intake will average around 60 degrees winter, 80 summer. Exhaust will be vented out to attic.

Very decent temps for a heat pump.

There are only 2 of us here the vast majority of the time. Guessing, I would say most of our electric bill is the water heater, especially after I installed the Emporia monitoring system, I can see that but never really kept data. My lack of knowledge at this point plays into a question of 1. How much power will this water heater actually use with the 2 of us?

About 1/4 as much as an equivalent standard dual element 50 gallon.

And 2. Is it even worthwhile to connect to my solar system?

Given that it's only two of you, and you're conscious of the implications of heavy hot water use and modify your behavior accordingly, then most likely.

It's worth checking that specific unit if you can find the owner's manual. Some have a selectable "mode" that will NOT run the element even with heavy use - your hot water is just let hot.

As my original premise is to have a home generator not requiring refuelling from outside sources do I add this to my system. You made the most pertinent point that adding a second inverter my idle current draw alone is significant. I'm thinking about installing a second 5kw Growatt in parallel for backup in case of equipment failure AND that would give me the ability to experiment. I want to get some running experience before I start adding. Tomorrow I will rewire my panels per your instructions, thank you again. I'm waiting for an Overkill brand 48volt BMS, backordered.

That would be viable. It would burn nothing while off.

5kW is a pretty decent amount, so if you can keep the water heater from hammering its 4500W element, the single unit might suffice.

Another area I neglected was to compute the solar required to keep my batteries charged, my home load and how much battery I actually need.

You aren't the first. If you manage to scrap less than $750 or so of stuff that you had to replace due to bad planning, you beat me.

Essentially I bought what I "thought" would be right. Another issue you and another person pointed out is the shading of my panels. 10 weeks there is a shadow from the fence, average 20% of the physical panel. I humbly admit I don't know the consequences of this. The panels don't shut down, oddly, at least to me, the panels generate power as soon as the sky brightens up in the morning and do very well on cloudy days. I did not know this. If there is a method of measuring this performance and associated loss I would appreciate that direction.

Impact of partial shading depends on the physical configuration of the panel cells.

This video shows the difference between horizontal and vertical shading on a standard panel:


What bi-facial do you have?
 

As I understand it, your panels are 2X 220W 72 cell panels in parallel to produce a 440W panel with a total of 144 cells.

In the vertical orientation, by shading an entire bottom row of panels, you're taking out the entire bottom half of the panel as demonstrated in that video.

If you were to stack the two panels horizontally, the top panel would get NO shading at all, and the bottom panel would have the lower 20% of each of the panel halves shaded, thus you would take out 1/3 of the entire panel vs. 1/2.

So in summary:
Vertical panels lose 50% due to shading the bottom row of cells, so entire array is reduced by 1/2 as production on the bottom half of all panels is vastly reduced.
Horizontal panels stacked two high, top panel has no shading, and bottom panel loses 1/3 on both halves, thus entire array is only reduced by 1/6.

I'm guessing on the routings of the cells, so you would need to conduct your own experiment or at least visually verify how the cells are connected.
 
Then this makes it worth it!
So when I switched my panels to 5s2p I had to move my ground wire for the panels and have voltage there? So I am using 2 panels that were damaged during shipping. Initially they tested fine but am wondering if that is causing this?
 
Note from:


It lists a 4500W heating element.

A heat pump is hard pressed to push water to 130°F for water heaters, but it's very effective for getting it most of the way. If you can be patient, it can eventually get you there; however, ambient conditions (cold heat sink) and demand may exceed its ability to perform to customer expectations, so it has a backup element.
So I've got the water heater installed and running. 1.34 amps in heat pump mode. Thanks again. This will easily work in a pinch. I have a lockout breaker so I can switch power sources
 
Looks like the fence would shade the panels some hours and seasons.
How about securing the lower end near the top of the fence?
And the upper end out far enough that the eaves won't shade it (maybe hang from the eaves)?

A priority switching scheme to avoid multiple heavy loads at once should let you go farther with less inverter. You might get to run those heavier loads.
SoC based control might avoid wasting available power when battery is full, and avoid draining it too far.
For motor loads, starting surge will be the limiter. Do you know or can you measure inrush? I assume 5x nameplate current, but also got a cheap meter with inrush recently (previously used oscilloscope and current transformer.)
Yes I can measure that, plus I have an Emporia VUE monitor, so I can track all my loads and startup inrush. I bought a second 5kw Growatt inverter, mainly for backup due to failure. But I am wiring it into my system so I can experiment
 
Another thought on the vertical panels that I didn't see mentioned is that I believe some manufacturers don't permit supporting at either end on their longer (>2m) panels (for surviving high wind loads I guess). They'd be sheltered by the fence somewhat but it might be worth checking the spec sheet.
 
Back
Top