diy solar

diy solar

First build, some nudging in the right direction would be a relief ?

LevelChamp8

New Member
Joined
Nov 14, 2021
Messages
8
It’s the end of 2021 and looking for the simplest set-up for my objective. To cut power bill at least in half for three 1000W appliances that run 24/7.

Looking to start out with:

5-10 ~400WATT Bifacial solar panels.

MOES Dual power controller switch

Either a Renogy or Growatt all in one unit (not sure if you need batteries are required for a particular Growatt unit or for what I looking to achieve)

Lithium Iron Phosphate batteries (if needed)

The cables and extensions (not sure exactly what I need yet, slightly overwhelmed by all the different approaches)

ANY and all sound help will not go unrecognized! Thank you, it means more than you know and this community seems awesome from what I’ve checked out!
 
I would forgo the Chinese products and stick with an Outback, SMA, or other non-Chinese system.. but that's just my opinion.

I'm assuming you want your setup to run off of battery power as much as possible until the bats are depleted, then the transfer switches flips back over to grid right?

That should work fine so long as the appliances you're running are not sensitive to power fluctuations. You'll also need to make sure you have some device that can read the voltage of the battery system and open or close a relay to activate the transfer switch when the battery bank gets too low.
There's a big difference between a 1000 watt resistive load and a 1000 watt inductive (motors) load... Resistive loads are easy to run, not so much for inductive loads.
 
Did you say you consume 72 kWh/day?
Depending on your utility rates, cutting bill in half might mean generating 36 kWh/day, or perhaps less is needed.
36 kWh average year-round typically means about 7kW of PV. Check an insolation calculator for your location.

Batteries will cost more than utility rates, unless you DIY with low cost cells.

Is net metering available? That is the most cost-effective. Depending on your rates, might pay back in 3 years.
 
I would forgo the Chinese products and stick with an Outback, SMA, or other non-Chinese system.. but that's just my opinion.

I'm assuming you want your setup to run off of battery power as much as possible until the bats are depleted, then the transfer switches flips back over to grid right?

That should work fine so long as the appliances you're running are not sensitive to power fluctuations. You'll also need to make sure you have some device that can read the voltage of the battery system and open or close a relay to activate the transfer switch when the battery bank gets too low.
There's a big difference between a 1000 watt resistive load and a 1000 watt inductive (motors) load... Resistive loads are easy to run, not so much for inductive loads.
Just did a quick lookup. Seems that it’s inductive being they are servers like computers that run fans as well.

It seems you are really are a solar enthusiast and I can already tell you know A LOT so I appreciate your reply!


If I can afford the price tag I will, given this is my first build I’d like to safely just get it up and running ASAP.
 
Did you say you consume 72 kWh/day?
Depending on your utility rates, cutting bill in half might mean generating 36 kWh/day, or perhaps less is needed.
36 kWh average year-round typically means about 7kW of PV. Check an insolation calculator for your location.

Batteries will cost more than utility rates, unless you DIY with low cost cells.

Is net metering available? That is the most cost-effective. Depending on your rates, might pay back in 3 years.
Need to look into net metering only heard the terminology a couple of times.

Safely we’ll say 84 kWh/day

I used 712 kWh in only 20 days running just one server. Which cost 114$ said and done

Say starting out with a smaller system I order 5 410W bifacial panels, a dual power controller switch, cables, extensions, mppt, inverter, charge controller and batteries. Other than basics, I’m wondering where to get sound info on set up safely and properly..total beginner status.

I’ll check out a calculator, appreciate the tip!
 
Net metering is where you spin the meter backwards and use the grid as if it were a battery. In some locations, buy/sell price is the same, some it is different. In some you are forced onto a different rate schedule (time of use) that may or may not be advantageous.

$114/712kWh = $0.16/kWh
That's at or below lowest tier where I am.

You can DIY a GT PV system for $1/watt of materials (assume labor free), and that will cost $0.05/kWh over 10 years, $0.025 to $0.03/kWh over 20 years depending on whether inverter needs replacing. So your payback could be 3 years.

Batteries wear out, so consider cost over lifespan. They range from a low of $0.05/kWh for DIY LiFePO4 to a high of $0.50/kWh for brand name lithium systems or for AGM lead-acid. I don't believe you can save money storing power in batteries (at least not at your rates.) Even for me where peak time rate is 2x or 3x off-peak, I find it more economical to export 2x or 3x the wattage from PV during the day to get a credit for what I use in the evening (refer back to GT PV costs vs. battery costs.)

Some hybrid inverters are relatively low priced and are batteries optional or can work with a small battery. Then you get backup for critical loads.

What I use is SMA, so I would consider the new model Sunny Boy 7.7kW inverters (a bit under $2000) if I needed latest features for permitting. If UL-1741-SA isn't required, I'd use older models than can be had cheaper. The new models offer "Secure Power", up to 2000W 120VAC manually enabled during power failures (PV direct to AC, no battery).
 
Just did a quick lookup. Seems that it’s inductive being they are servers like computers that run fans as well.

It seems you are really are a solar enthusiast and I can already tell you know A LOT so I appreciate your reply!


If I can afford the price tag I will, given this is my first build I’d like to safely just get it up and running ASAP.
Sounds like bitcoin mining? That would be more of a resistive load even though there are cooling fans.

Inductive loads are like motors that run air compressors, air conditioners, water pumps, or very large fans. While computer fans, desk fans, the fan on a wood stove, etc, are technically inductive loads, they are very small in nature.
 
Computer power supplies? That would be a rectifier/capacitor front end with poor power factor, unless they have active power factor correction (in which case they should look resistive.)
Some newer Dell computers do, and require sine wave input.
 
So gentlemen, between the two your apt minds my toddler level comprehension in this field begins to weigh down on me.

Without being a wasteful spender, especially in this climate, what should my next step be. I’d like to think I can order the panels, get a list of specific units needed to utilize the panels, run the mining hardware off the array and when the power demand isn’t met it switches to grid power...now for the flaws in my thinking..
 
So gentlemen, between the two your apt minds my toddler level comprehension in this field begins to weigh down on me.

Without being a wasteful spender, especially in this climate, what should my next step be. I’d like to think I can order the panels, get a list of specific units needed to utilize the panels, run the mining hardware off the array and when the power demand isn’t met it switches to grid power...now for the flaws in my thinking..
Next steps:
1) Find out what your net metering rules are because that's usually the best way to go. As Hedges said, the grid is your battery.
2) Find out how much juice you're actually using and how you're using it.

Grid tie is the best way to go.. usually. Even in states with poor net metering rules, it is still usually the best way to go..

Batteries are ridiculously expensive and they are rarely an economic solution to anything but a power outage or when the grid is not available.

I have a friend who just installed a non-permitted grid tied system.. Illegal, but its as safe as any other system. Basically, he monitored his power usage in his house and found that he is never drawing less than 800 watts.. Ever.. and usually he's up in the 1500 to 2000 watt range.

So what he did was to install two 400 watt solar panels with micro-inverters.. When the sun shines, he saves money How much money he saves depends on how much the sun shines, but unlike legally permitted grid tied systems that are intentionally oversized to make up for the night time power draw, he doesn't save as much..

I'd recommend you go with a grid tied solar system.. Contact your power company, fill out the paper work and file the proper permits.. then cover your house or lawn with solar panels.
 
It is difficult and expensive to build a PV/battery system to power large loads.
Having grid available to switch to does help.

"Net metering" is ideal, but requires some cooperation and permission.
"Zero export" is slightly fancier equipment but goes unobserved by the grid operator.
Some inverters will use current transformers to measure grid connection.
Some just measure current coming from grid into them, and use PV as priority to the extent possible.

Given a high fixed load, if you know it will never drop below some particular wattage, you could connect a grid-tie UL-1741 compliant inverter. It would put power into your house, but the mining hardware draws even more. I would have PV arrays aimed and morning and at afternoon sun, producing power for much of the day.

Pretty much what Murphy just said (but watch out for Murphy's Law)
 
One question no one has asked is where abouts are you located? If you're in Las Vegas, it might be reasonable. If you're somewhere like here in Oregon, or England, or quite a few other places, the numbers get silly. Where I live, to generate that 36 KWH per day in the winter, you'd need 30 KW of panels, or more. The sun is low and there's lots of clouds for 4-5 months per yer. My 10KW of panels generated 9.3 KWH of power today, which is pretty typical this time of year. You should go to an online solar irradiance calculator and see what kind of production you can expect.

Panels are cheap these days, but the inverters and charge controllers are still expensive if you want quality, and in my experience it's all the little things that add up to a much bigger number than you expect. I just added 5KW of panels and another charge controller and spent about $450 on wire, crimp lugs, cables and such, $150 for load panels and breakers, $100 for fuse holders and fuses, etc. Honestly, if you're just looking at it in terms of saving money, it's going to be hard to compete with the power company unless you're somewhere with extremely expensive power.
 
Northeastern Indiana. Skies are mostly clear. And in the winter there’s generally some snow on the ground so added benefits of bifacial panels.

Likely will be running about 4100W 24/7 once everything is up and running.

To keep it most simple, say I’ve got 5 410W bifacial panels in an array in the yard. Let’s say I’m not doing net metering.

What is the best way to utilize this array, including the components I need, in order to use the PV until it gets dark and/or the battery system is depleted and can instantly switch to grid power from there until the next day when the sun is up and the solar system is outputting again...

Thank you all for responses and insight, doing my best to take it into consideration while also seeing which route is actually viable.

I moved from city to rural and commercial electric to utility bill so gotta get something going quick cos the power cost is actually close to a break even at the moment.
 
IMHO the first thing to do is determine what level of municipal signoff is required.

if your required to have permits, you best start with what the local municipality will accept for hardware. I dont know of any locations that will allow grid tie on non-UL listed equiment.

if your confident in your non-permit required status the quickest way to be up and running and cutting power cost is to use growatt Spf 3000tl lvm. installing a 120v only critical loads pannel and have the growatt power the critical loads. With a couple of used car batteries you can be harvesting sun while its up. As budget allows you can add more PV, a second inverter (or more) and move to a real battery.
 
moved from city to rural and commercial electric to utility bill so gotta get something going quick cos the power cost is actually close to a break even at the moment.
You may not save any money for a while with the purchase price of solar…
Not discouraging, just sayin’
 
For bootleg system, I would buy an old grid-tie inverter. SMA, Fronius, etc.
Hook it up, and with 2000W of PV but 4000W of load, there won't be any backfeed.

Sometimes utility company observes the panels and will give you trouble.

If grid tie net metering is available, that may be the best way to go. Maybe not if the force you on to a time of use rate schedule.
With time of use, it may be better to add 25% to 33% more mining hardware and only operate during off-peak hours.

You may not save any money for a while with the purchase price of solar…
Not discouraging, just sayin’

Decent new hardware (and labor free), I estimate 3 year payback.
Used panels and inverters would shorten that. Then, mounting hardware can dominate cost.
 
For bootleg system, I would buy an old grid-tie inverter. SMA, Fronius, etc.
Hook it up, and with 2000W of PV but 4000W of load, there won't be any backfeed.

Sometimes utility company observes the panels and will give you trouble.

If grid tie net metering is available, that may be the best way to go. Maybe not if the force you on to a time of use rate schedule.
With time of use, it may be better to add 25% to 33% more mining hardware and only operate during off-peak hours.



Decent new hardware (and labor free), I estimate 3 year payback.
Used panels and inverters would shorten that. Then, mounting hardware can dominate cost.
Okay so these are the specs of some poly panels I can get 10 of instead of the bifacials.

I looked up the SMA grid tie inverter. Wouldn’t be exactly sure which to get and if that’s all I’d need for a basic set up that switches to grid power after sun is down and PV output is too low for my load.

The pic is the spot I’ll put the array, took at about 14:00 today...
 

Attachments

  • 7B399DDF-E5D9-4978-BD1C-4B8629FB3FC5.jpeg
    7B399DDF-E5D9-4978-BD1C-4B8629FB3FC5.jpeg
    170.8 KB · Views: 4
  • F8D52F25-E860-4D64-92CC-2B3089D7DBFD.jpeg
    F8D52F25-E860-4D64-92CC-2B3089D7DBFD.jpeg
    274.4 KB · Views: 4
Something like this. I bought a pallet of them being closed out by a dealer for about $500 each ($0.10/watt)
They come with a fused DC disconnect. Connect 1 to 4 PV strings to the PV input, and connect the AC terminals to a 2-pole breaker for 240V (or 208V or one pole for 277V)
Turn it on and it works. (It auto-detects voltage if neutral is wired and jumper is in default position. Other jumper positions to set voltage if no neutral. Default wired for negative ground PV but jumperable for positive ground.)


No switching between inverter and grid. Your loads connect to grid, and inverter connects to grid trying to backfeed it. But if your loads exceed inverter wattage (or panel wattage), nothing ever backfeeds. With a data interface and PC or other host I think you can alter its maximum output power if needed.
 
Okay so these are the specs of some poly panels I can get 10 of instead of the bifacials.

37.6 Voc x 10 panels in series = 376 Voc at 25 degree C. Add 20% for conservative estimate in cold, 451 Voc max. OK as one string into a 600V max inverter.

30.3 Vmp x 10 = 303V nominal operating. Goes up and down a bit with temperature and illumination.

Inverter spec: Min. DC voltage / start voltage 250 V / 300 V

Start voltage of 300V is fine, 376 Voc minus 20% if really hot would still be 300V.
Minimum operating voltage of 250V is 17.5% below 303V nominal. Temperature won't get it that low. Low light maybe. But that's OK, it will still operate, just won't go below 250V looking for maximum power point.

250W x 10 = 2500W STC, somewhat less PTC.
If you could get 20, 30, 36 panels, they would work with this inverter in 2 or 3 strings. (multiple orientations for 3 strings.)
You can mix and match panel models, but each string should be same model or same current, and all strings should be same voltage.
 
Back
Top