diy solar

diy solar

Newbie - multiple arrays in different locations. Different sizes? Inverter/battery locations? Microinverters vs optimizers?

67King

New Member
Joined
Apr 23, 2023
Messages
39
Location
Friendsville, TN
Building a house, breaking ground on June 1. In east Tennessee, near Knoxville. Have given up on solar shingles, but the good news is that we have space and a few locations where we can put arrays. We are too early to have any sort of loading numbers, so the only thing I have to go by is benchmarking out current house. This one will be abotu 20% larger, but there are some major differences. First, it is on the lake, so we'll have a lot more windows. That and the sizing will increase the loading requirements. Our average electricity usage here is 2250kw-hr. We have some gas appliances that add 77 therms. That direct calculation doubles the loading; however, when I benchmark electric vs gas, the 29.31kw-hr/therm doesn't really hold. So if I use a modified version, I add a bit under 1000kw-hrs. So this house would have a fully electric load of call it 3250kw-hr per month on average.

A big unknown at this point is if I'll put in a lap pool. Really hoping to, as we're a big swim family (all 3 kids swim competitively, one is a D1 swimmer, and I do, as well). That would stand to take a lot more juice, especially if I heat the pool for a few months out of the year.

WOrking the other way. Going to try to avoid gas if at all possible. Natural gas utilities do not support the area, so we'd be looking at a buried LNG tank. That stuff is about 4 times more expensive than utility provided. We wanted to use induction cooktops anyway. Given the site, geothermal HVAC is ideal, so by going with that, we are fine with a straight up heat pump, rather than a dual fuel system (well, multiple - 2 level plus walkout basement home, a bit under 5000SF of heated space). That'll be a huge step in the right direction, but at this point, I cannot quantify it. Tankless water heaters will be another. Going with a metal roof with a radiant heat barrier. Will used closed cell foam insulation in strategic parts of the house. More generally speaking, this house will be newer, nicer, and much better built than the benchmark one. The plan here is for this to be our last home, so we will spend the money upfront to have a positive rate of return for >30 years (well.....hopefully, we'll both be 50 when we move in).

Main array will be very close to the house - if not for the terrain, I'd build an awning attached to the house. Will face SSE. Will use it as a "carport" to park a truck and 28' enclosed trailer. The length of that side of the house is 42 feet, so I have been limiting myself to that length. The goal is for the array to somewhat look like an awning against the house when viewed from the side (note that I'm going to space it off of the house by a few feet given the slope of the hill on that side, so I can have a step like feature, wall or not, then grade out a flat spot that would be appropriate for parking the truck and trailer). May be able to go longer, but that's the starting point. Structural support beams spanning 16 feet are kind of the end of the spectrum. So getting somewhere near a 42X20' covered area (accounting for overhang - and yes I may need to have some support for the overhanging portion of the panels if I go with longer ones).

So for now, I'm assuming I will have a 12X3 array. Best case is larger panels, something like a BlueSun460, which would give me a 12X3 array with a 16.5KW system. Something like a 400W Hanwha will also give me a 12X3, but a slightly lower capacity 14.4KW system.

That's teh straightforward part. Main panel, going to the house, use optimizers to get DC to an inverter in the house, supplying about 3 batteries (best case, Tesla Powerwall+ system, if those jackholes - and I drive a Tesla - ever let me buy the system, rather than forcing me to pay a 100% markup by going through a supplier in this area). Well......I *think* that's the straightforward part, anyway.

Here is where teh questions come in. Let me start out by saying we aren't going to mount on the roof of the house. However, there are two other areas where mounting panels will be an option. One is that I'm going to build a shed, about 12x16. It will have a 12/4 pitch, so I can probably get good use out of both sides. That's a pair of 4X1 arrays, giving me an additional 3200-3680KW. I am not entirely sure where on the property this will be. May be by the dock, may be far from it, depends on what TVA has to say, I suppose. Ideally for what we are using it for, it will be oriented so the panels would face ESE/WNW.

Boathouse/dock. I'm not sure how big that will be. There are piers there now, which appear to be abotu 30' wide (I can't recall the exact width, I will try to update this when I measure them). Probably 38' deep, as well. They, too, would have panels facing ESE/WNW. COuld have an 11X2 array on each side, giving a total of 17-20KW. Do need to point out that the horizon is lower on the WNW side than the ESE side, so I'm not sure which direction would see more sunlight.

Anyway, so between the 3 places, I'm assuming I should ahve enough to have a completely self sustaining system. Still likely will tied to grid, but just for convenience, not necessity.

What's the best way to handle this, in y'all's opinion? Main grid, like I said, appears easy. Optimizers to keep DC to send to batteries, then inverter, then one of the main panels (will have two 200A panels). Given the run required from the boathouse to the main panels, should I plan on a battery(ies) there, inverter, then send AC to the house? Or should I just plan on having that setup be a supplemental AC power supply to tie directly to the panel without battery backup? Figure I could send a pair of cables, maybe one for each panel, to a 50A breaker in each box, to supply juice only while sun is out? I am assuming since I don't own stock in any copper mines that I'd want to avoid running DC over such a long distance. Guess I could also put the more critical components on one of the panels, and the less critical on another, and have all my battery backup for the primary panel, and the supplemental to the secondary?

Is there such a thing as a rectifier in solar? Meaning could I run microinverters on the boathouse/shed arrays, and send AC to the house, into a rectifier, and then merge it with the DC source from the main array going into battery storage?

I know that's a whole lot going on. I'm trying to figure out the best strategy. I have time to sort out all fo this, but most of what I find out there is for more conventional/straightforward systems.
 
Bump.....anyone? Just trying to lay out a strategy at this point, not details. I'm at the "basics" point, here. I'm an engineer, but NOT an electrical one, so I'm pretty analytical, but don't know the details. A bit lost with all of the acronyms in conjunction with brand names and models used here, so apologies if my questions seem overly basic (i.e. not claiming to be an expert, but I'm not a complete fool).

Biggest question I have, really, is how to move power from a source a couple hundred feet away to the house. That distance makes AC the better choice, but there's not really a good way to put a battery bank and inverter near the panels (that and frankly, should have mentioned previously, I'm not sure TVA would permit that kind of equipment to be installed where housing structures are not permitted).
 
Grid tied definitely. Cheaper than generator, repaired by crew.
What are rates? Is net metering available?

Gas is good for heat and backup power during storms and outages.
PV is good when the sun shines. Multiple orientations flattens production curve.
I favor oversize PV, undersize battery.

Long runs of high voltage DC from PV is good. Losses higher when production is higher, voltage drop doesn't affect functionality (as it can for AC), is only loss.

I use Sunny Boy and Sunny Island for split-phase grid backup. It may or may not have the best features for zero-export if no net metering.
I might consider 3-phase setup if mostly off-grid, never planning to draw more than 6.7kW (limit of split phase feeding into one out of 3 inverters in 3-phase system.) 3 phase breaker panels cost a lot more, but would be good for pumps and tools.


Schneider is another reliable low-frequency inverter, generally DC coupled but supports AC as well.

Yes, you can have AC from inverters later charge batteries. That's called AC coupling. I do that with Sunny Boy string inverters but it works with microinverters as well. They should support "frequency-watts", part of UL-1741SA or SB.

Tankless water heater? Why? Use a storage tank water heater; that's extra energy storage. Maybe heat pump, maybe gas, maybe resistive. A PV system should be sized to generate surplus, and resistive loads can be a dump load. In your situation I would suggest having both gas and electric available, maybe cascaded water heaters.

The most cost effective batteries appear to be LiFePO4 server rack batteries. Need enough to meet peak surge current. Select battery and inverter so they communicate.
 
Grid tied definitely. Cheaper than generator, repaired by crew.
What are rates? Is net metering available?

Gas is good for heat and backup power during storms and outages.
PV is good when the sun shines. Multiple orientations flattens production curve.
I favor oversize PV, undersize battery.

Long runs of high voltage DC from PV is good. Losses higher when production is higher, voltage drop doesn't affect functionality (as it can for AC), is only loss.

I use Sunny Boy and Sunny Island for split-phase grid backup. It may or may not have the best features for zero-export if no net metering.
I might consider 3-phase setup if mostly off-grid, never planning to draw more than 6.7kW (limit of split phase feeding into one out of 3 inverters in 3-phase system.) 3 phase breaker panels cost a lot more, but would be good for pumps and tools.


Schneider is another reliable low-frequency inverter, generally DC coupled but supports AC as well.

Yes, you can have AC from inverters later charge batteries. That's called AC coupling. I do that with Sunny Boy string inverters but it works with microinverters as well. They should support "frequency-watts", part of UL-1741SA or SB.

Tankless water heater? Why? Use a storage tank water heater; that's extra energy storage. Maybe heat pump, maybe gas, maybe resistive. A PV system should be sized to generate surplus, and resistive loads can be a dump load. In your situation I would suggest having both gas and electric available, maybe cascaded water heaters.

The most cost effective batteries appear to be LiFePO4 server rack batteries. Need enough to meet peak surge current. Select battery and inverter so they communicate.

My goal is to be self reliant power wise. Net metering isn't offered, so the only real reasons I mention being grid tied are to have power until i get it all sorted, and in the event I'm a little short. The provider lists costs on their site at $0.105, but my bills have been closer to $0.13, so either they have not updated their published rates, or the fixed charges on my bill are much more sizable than I had realized, and they have gone up. Do not believe the bill breaks down fixed vs variable charges, but I don't have that in front of me, now, as I'm away from home for the week. And even though power is cheap, I still plan on living in this house the rest of my life (I'll be 50 when we move in), so I still should expect a positive NPV.

One of the ways I'm hoping to be entirely energy dependent is by going with the more efficient systems. And fortunately, a geothermal system means i don't need the dual fuel setup for my heat pump that I otherwise would. And like I said, natural gas isn't an option, I'm stuck with LNG, which is about four times more expensive. It also means that the more efficient tankless setups are going to be my going in assumption.

The AC coupling thing is EXACTLY what I was looking for. Thank you! I didn't realize that several of the battery systems (including the Tesla I referenced) were that. That clears up why I keep seeing microinverters discussed even when battery storage is in the system, as well. I just thought I was confused.

I assume the LFP racking system you mention is a DC coupled system? Trying to paint a 30,000 foot view, going off of the simpler terms I know (i.e. panels and batteries are DC, inverters change DC to AC, and the only way to put stuff in a battery is DC). So it almost sounds like I may end up with two stand alone systems. The one close to the house could be a DC coupled with racks, and the second one where the panels will be mounted far away would be an AC coupled. I expect to have two 200 amp panels.

I wonder, though. If my DC system gets fully charged, could I tie it over to my AC system in the event it is not?
 
My goal is to be self reliant power wise. Net metering isn't offered, so the only real reasons I mention being grid tied are to have power until i get it all sorted, and in the event I'm a little short. The provider lists costs on their site at $0.105, but my bills have been closer to $0.13

DIY GT PV (or AC coupled portion) costs $0.025/kWh, turn-key installed costs $0.10
Battery and battery inverter costs more, so I prefer to minimize it.

Generator costs $1.00 so only for emergencies.

Although net metering isn't offered, will utility permit you to backfeed for zero credit? If so, you don't need the added complexity of zero backfeed. An AC coupled grid-backup system will charge batteries, then export surplus to grid. When grid is down it varies island grid frequency to curtail unneeded power.


I'm stuck with LNG, which is about four times more expensive.

On the bright side, it keeps forever. Unlike gas or diesel.

It also means that the more efficient tankless setups are going to be my going in assumption.

"More efficient" in and of itself is an assumption.
I've seen nothing compelling or even persuasive about them.

Surplus PV you have otherwise curtailed is the lowest cost power. A tank type heater lets you capture and store it.

The AC coupling thing is EXACTLY what I was looking for. Thank you! I didn't realize that several of the battery systems (including the Tesla I referenced) were that. That clears up why I keep seeing microinverters discussed even when battery storage is in the system, as well. I just thought I was confused.

I assume the LFP racking system you mention is a DC coupled system?

No, it is a battery. It could be connected to SolArk, Schneider, Midnight Rosie, offering various combinations of AC and DC coupling. For Sunny Island, look for a brand it talks to (list on SMA's website), might include BYD. For DIY batteries REC is compatible BMS. SMA primarily AC but also supports DC coupling.

Trying to paint a 30,000 foot view, going off of the simpler terms I know (i.e. panels and batteries are DC, inverters change DC to AC, and the only way to put stuff in a battery is DC). So it almost sounds like I may end up with two stand alone systems. The one close to the house could be a DC coupled with racks, and the second one where the panels will be mounted far away would be an AC coupled. I expect to have two 200 amp panels.

I wonder, though. If my DC system gets fully charged, could I tie it over to my AC system in the event it is not?

Try to make it one system, but with redundancies and load shedding of less important loads. You can make a system with DC coupling and also AC coupling through the wires going to other buildings. Put in some signal wires so controller can shed loads at the remote locations while keeping AC coupled PV connected.

Start with power audit including kWh/day by season, max continuous kW, peak kWh starting surge. That will narrow what hardware meets the needs.

Midnight Rosie is worth looking at. HF inverter with long surge, from the guys who invented the industry and designed half the big names. But otherwise, people have traditionally used LF for heavy loads. Nothing intrinsic in the architecture (except early Trace inverters), more a matter of how over-built they are. Many people seem happy with HF SolArk.

I would be inclined toward 3x Sunny Island (18kW) coupled with 36kW of Sunny Boy. (I think I've got that on my shelf already, a couple times over ? . Just not enough room for all those panels nor big enough loads to need it.)
 
My goal is to be self reliant power wise. Net metering isn't offered, so the only real reasons I mention being grid tied are to have power until i get it all sorted, and in the event I'm a little short...................
I was actually mistaken. Apparently TVA backed out of the plan to eliminate net metering, but found some info about it this morning. That simultaneously alleviates and complicates things. Can probably somewhat follow this guy's approach, but at a smaller scale: https://diysolarforum.com/threads/my-diy-60-kw-220-panel-setup.46585/ (though without the bitcoin mining stuff he has going on that uses so much electricity).

Seems that a pair of EG4 systems are probably my best bet. Was thinking I'd have two separate 200A panels, isolated from each other, unless I could tie together at battery banks. Was thinking one being off-grid, and the other grid tied with a hybrid inverter. Main system looks more and more like a 12X3 array of BlueSun 460W panels. Secondary system can fit a 10X3 of the same array with an ESE orientation, and I could put several more (up to the same array), but they'd be facing WNW. I guess, though, which I hadn't thought of before, I could wire it all in series, and have a low current, high voltage DC electricity running the couple hundred feet to avoid heavy gauge wire, then run to the inverter. May still want optimizers for the array that will have some panels facing a poor direction.

My builder likes to keep the HVAC on as soon as drywall goes up, so I may want to get one of the systems up and running before I have the remainder of the appliances ID'd, so I won't be able to have a
 
Back
Top