diy solar

diy solar

30kW "cheap" inverter

BlueSmoke

New Member
Joined
Feb 4, 2021
Messages
16
I'm planning the utilities for the off grid home I'm in the process of building. We will be building in central Utah USA. The goal is to live a normal life and not have to worry about conserving too much, but be off grid. There's good sun in the summer, and regular sun in the winter, but there's a few day string each month in the winter where very little PV power can be generated. We are most likely going to be doing a combination of PV+battery, generator, and propane. I'm sizing loads currently and it's looking like I'll need 24-30kW of inverter AC output to the house. Imagining a worse case scenario, it's Thanksgiving or Christmas, lots of family is in town, people are showering, cooking, etc and it's snowing heavily. So I need that peak 30kW power to be available from the batteries... Which means a big inverter system.

I've been learning a bunch about how to stack inverters and wow does it get expensive to hit those bigger power numbers. Does anyone know anything about these inverters? http://energetechsolar.com/30kw-sdp-off-grid-inverter They are sold in a bunch of places. They aren't all that efficient being 93-94%, I don't think there's a charger integrated in it, and there are quite a few other features that needs to be added which are normally included in an all-in-one hybrid inverter system. Has anyone looked at those bigger inverters? I haven't found much online except that it looks like they are Chinese manufactured. They are big a cheap though, so it makes me wonder what's wrong with them other than not being full featured.

In doing a large install (for residential anyway) like this, is there any downside to using more discrete components? Like instead of using a bunch of hybrid inverters, using a built up system of PV controllers, inverters, battery chargers, generator auto start/shutdown, etc.

In the end we'll probably end up with more propane appliances to limit the amount of solar we need due to the high cost of the inverters and battery storage, but I want to fully explore this before making that choice. Who are the forum members that have BIG installs that I can check out?
 
24kW to 30kW for an off-grid house is relatively ridiculous, but not impossible.
My picture shows 4x SI 6048US, which can produce 23kW continuous at 25 degrees C, 28 kW for 30 minutes (about what you want)
More reasonable would be 2x SI 6048US for 11.5 kW from battery + 4x 6kW Sunny Boy for 24 kW from PV while the sun shines.

I think you're better off prioritizing loads, e.g. switch off one heavy load (furnace?) when running another (oven?) so peak power isn't as high.

How many kWh/day to you anticipate?
What battery do you plan to use?
A diesel or propane generator could support periods of higher power draw, more cost-effectively than a huge battery.
 
In our current 2000sqft home we use about 20-25kW per day. Estimating for the new house, it's looking like 90kW for a high usage day isn't out of the question. I expect something closer to 30-45kW most days. The house will be larger, 6000 sqft with about 1000 sqft unfinished storage, plus a shop with 240v tools. We will be using lots of insulation and paying attention to air sealing, so there shouldn't be a huge amount of cooling and heating load required. Our current home is not highly insulated or really well sealed. It isn't horrible, but it isn't great either. We have natural gas doing most of the heavy lifting today with the furnace, water heater, and stove/oven. Our hope is that the extra insulation and air sealing will significantly reduce the need for heating and cooling along with taking advantage of passive solar heating options.

I am also thinking that prioritizing loads will be in the future if we go for more reliance on PV, but I want to avoid it as much as possible to keep my wife's life simple. Since most days are pretty sunny, and we only use lots of power when we are awake, I'm hoping to oversize the panel array and work on more of a use it or lose it lifestyle. We have plenty of property for as large of a ground mounted array as we like.

Right now I'm looking at the Tesla batteries. I can't get their full capacity with SMA (or any other manufacturer's) gear due to the weird voltage, but they are so cheap I can just buy more to deal with the voltage issue. Though, I just did the battery math on my proposed holiday crazy usage situation and it won't work. I'm planning to start with a 10 Tesla battery module array and assuming I can use 3kW of the 5kW available in each module. So I'd only have 30kW battery available at full capacity which would last for a few hours of insanity (since it won't actually be 24-30kW constant) and then the party's over. I'll absolutely have to plan to have a decent sized generator to handle the loads during excessively high usage and on zero PV days which could correspond with high usage days. And it seems like I'll have to plan a larger battery bank too. Plus think about moving some appliances to propane.

Seems like my dream of limiting propane to a gas cooktop and small generator are dead via math lol.
 
You should plan on using the generator for the peak period during the winter holidays, otherwise you will be dramatically oversized for the rest of the year. Take a look at where your power is going to make some informed decisions.

I'll throw a lot of general ideas and tips towards you here and see what sticks. Some of it may be useful.

Some general tips to think about since you are building everything new:
  1. Buy the best windows you can get - you lose most of your heating/cooling to the outside through the windows
  2. Put in as much insulation as you can. Two-part polyurethane works wonders if that is a possibility, but tends to be more expensive than fiberglass batts
  3. Put in a huge hot water heater and heat it using extra power from the solar (more on this later)
  4. Build in some thermal mass in the house and look at heat pumps to see if they will work in your climate. Some of the new heat pumps will work down to very low temperatures
  5. Look at capturing heat from the generator if you can. You will probably be running the generator in the winter during snow storms which is perfect since you can capture that waste heat to heat your home.

House Design​

For the house, consider some passive-solar style design. You can greatly reduce your summer air conditioning if you have overhangs on any sun-facing windows (would be south-west facing for you up there in the Northern Hemisphere) and the winter sunlight will come through. The low-E windows will unfortunately bounce a lot of that heat out, but they are still worth it.

Hot Water​

You can use hot water to store that extra solar energy. Heating hot water takes a lot of power, but you can store it easily. If you use a heat pump to heat the water during the day, it is both warmer outside so the heatpump is more efficient. You can use the electric elements as a load dump. Heat pumps probably were not an option for you in the past, but the newer ones with vapour injection may work at the extreme temperature ranges you see.

HVAC​

If you use an air-to-water (or potentially a water-to-water) heat-pump, you can use that stored heat for both your hot water and heating your house. Take a look at hydronic air handlers. You will want to put a glycol mixture in the air handlers to prevent freezing if your heat is off.
 
I think the OP is not distinguishing between kW and kWhs. For starters that would inform the conversation if he would make that distinction. . Ten Tesla Powerwalls would be 135,000 kWhrs. The Powerwall is not designed to be used off grid as the primary inverter but perhaps what @BlueSmoke is referring to is grid zero which which simply means not using much grid power.

On the other hand if he is talking about 24 volt battery modules from a wrecked car then he may be able to put two in series and get to a nominal 48 volts that most inverters can deal with. I was considering using Tesla modules but found a less expensive option using LFP. I drive two Teslas so I have respect for the quality of the modules but I also know that without the cooling system in the car that those modules are much more dangerous. Also the extra cost of multiple BMS systems to monitor and parallel the modules ends up making that approach a lot more expensive than I was able to purchase 42kWh of LFP batteries for. My average cost not including my BMS was less than $125 per kWh. I have not found Tesla modules for less than $200 per kWh unless in mixed quantities. If one cannot get all the modules from the same vehicle, there are other challenges using mixed modules.
 
Last edited:
The numbers you have to come up with include:
Peak kW with many loads running at once.
Surge kW to start motors (typically 5x nameplate) such as A/C and well pump.
kWh/ (24 hour) day consumption you need to generate somehow. PV? Distinguish summer from winter.
kWh needed for a period without sunlight. Each night? Extended period of cloudiness?

Batteries are the most expensive component, when priced in terms of $/kWh capacity or lifespan. DIY lithium is bringing it down, but still twice the price of PV (meaning you spend 3x as much for a kWh saved then used at night, as a kWh used the moment it is produced.)

If your nighttime loads are minimal, that's all the battery you really need. If you want "3 days autonomy" running off stored power, battery needs to be much larger. Probably not just 6x larger, maybe 9x to 12x larger as for one night.

You can put in a modest battery and a larger battery inverter so the wife usually doesn't notice.
By running well pump and water heater only when there is surplus generation available, you store energy in something much cheaper than lead or lithium.
With excessive PV (AC coupled, or DC coupled if battery charge current kept within limits), you can "use it or lose it" as a cheaper solution than storing every kWh produced. PV panels are cheap!

I have about 24kW battery inverter, 12kW PV inverter, 20 kWh (14 kWh usable) AGM battery. That's for grid-backup (capacity for a single night, expect < 350 power outages in a decade).
Some people use forklift batteries for off-grid living, more capacity for a given price and possibly twice the lifespan to boot.
One thing lithium buys you is that you can discharge 80% every night and still get thousands of cycles. To me, less value if sized so large that it rarely gets deep-cycled. Good if you plan on 3560 deep cycles in a decade.
 
Ok guys, wow, this has been so helpful.

@HateGrid We're building an ICF rambler with basement with additional insulation everywhere. Extra south facing windows to capture the heat in the winter with shading for the summer. I'll look into advances in heat pumps. I wasn't considering them because I didn't think they would work in a true 4 season climate.

I didn't think about using hot water as a battery of sorts. We will be installing hydronic radiant floor heating, so dumping excess energy into a large insulated tank of hot water and using that for potable water + a heat exchanger for the floors is a really good idea. I was feeling bad about wasting a whole bunch of production by oversizing the array to deal with winter. I am already planning to have about 1 month well water storage, but that is just plain water. That's a great point that I'll only have to run the well pump when there's excess solar production.

@Ampster I'm talking about individual Tesla battery modules salvaged from wrecked vehicles. I didn't realize that LFP was getting that cheap now. I'm assuming you picked up some grade B cells at that price? Even with BMS and other parts that would be cheaper and more usable than my idea with the Tesla modules. Do you have a thread I can read about your battery bank?

@Hedges How much battery storage do I need? Enough to get through one bad solar day without hitting the generator I think. I was thinking that I wanted a full week of battery storage, but that's totally unrealistic for my situation. Diesel is a really cheap battery, and if I have to have a generator (it seems I do), it might as well get some use charging the battery bank on consecutive poor solar days. I just don't want it starting up every day in the winter.

I did a bit of reading last night about oversizing the array and reducing the battery bank size like you are suggesting. I found a paper out of Australia modeling oversizing the panel array and reducing the battery bank and looking at how many blackout days they would have based on solar irradiance. I've downloaded a bunch of PVOutput data from installations near where we will be building to do a similar study for that locality. That will help us decide how much power we can realistically make for a certain amount of initial investment. From there we will have to decide whether we want to scale back on the house or what the plan will be.
 
@Hedges How much battery storage do I need? Enough to get through one bad solar day without hitting the generator I think. I was thinking that I wanted a full week of battery storage, but that's totally unrealistic for my situation. Diesel is a really cheap battery, and if I have to have a generator (it seems I do), it might as well get some use charging the battery bank on consecutive poor solar days. I just don't want it starting up every day in the winter.

I did a bit of reading last night about oversizing the array and reducing the battery bank size like you are suggesting. I found a paper out of Australia modeling oversizing the panel array and reducing the battery bank and looking at how many blackout days they would have based on solar irradiance. I've downloaded a bunch of PVOutput data from installations near where we will be building to do a similar study for that locality. That will help us decide how much power we can realistically make for a certain amount of initial investment. From there we will have to decide whether we want to scale back on the house or what the plan will be.

The minimum battery you need is one night. That's what I have, but my application is grid-backup. If you did that, you would have to run generator more often.

Traditionally, people off-grid wanted 3 days worth of battery power. That was when PV panels cost 20x or more what they do today. It is reasonable to put in 4x the amount of PV you need on a good day. If you wanted 12.5kWh/day in the summer, 2500W of panels would do that, about $500 or so. You could put in 10 kW of panels for $2000, still not much compared to balance of system.

Ideally you'd have historical daily insolation data, could model how often generator was required. Sounds like you found that.

Heating is the big load. You can forgo laundry and well pump until the sun comes back out (given large enough closet and water tank). Other than heat (which might include wood?) you should be able to drastically cut consumption to stretch battery life.

If you do put in a generator, set it up for CHP, combined heat and power. If water cooled, before the radiator plumb to a liquid/liquid heat exchanger, and plumb the heated water from that to a pipe with exhaust pipe routed through the middle (appropriate materials, of course.) About 2/3 of your fossil fuel's energy is lost as heat, so capturing it heats your house.

The bigger a house is, the smaller it's surface area in proportion to volume. Assuming a boring sphere/cube rather than interesting architecture.
I think you should build what you want, don't let PV constrain you.
 
Sounds like you have a great house design. The passive-solar approach of getting sunlight in the winter and shade in the summer works wonders at reducing your HVAC load. Nice deciduous trees work well, but then you have to deal with leaves in autumn and the risk of branches breaking off in storms.

With the hydronic radiant floor heating, are you using a concrete slab? That adds excellent thermal mass which helps you store heat, so you could run the heat pump during the day when on solar and let it coast overnight. If your house is well insulation, this may very well be enough. Typical HVAC sizing guidelines can be thrown out when you have a lot of thermal mass since most HVAC assumes a small amount of thermal mass for timber construction and bringing the house to the target temperature in a few hours. It sounds like you are handy with analysis and with a new build, you know the R-value of the house, so you should be able to do your own analysis for heating and cooling to size your equipment.

If you have lots of open ground space, you can do a geo-source heat pump. A typical forced air heat pump is air-to-air. A hydronic heat pump is air-to-water (and then you have a fan-coil unit inside that does water-to-air). A geo-source heat pump is water-to-water and uses loops of pipe buried in the ground. This provides a better average temperature for the heat pump so it will be much more efficient. The only issue is the extra cost (and the people that do these systems tend to charge a premium). In the end, air-to-water is typically the best financial decision.
 
  • Like
Reactions: Dzl
If you're off grid, you're going to have a septic system, with trenches below the frost line, right?
How does the length/area of geo-source compare to leach field?
Not much left to spend just buying tubing, if you get the trenches for free.
 
How does the length/area of geo-source compare to leach field?
Not much left to spend just buying tubing, if you get the trenches for free.
Genius thinking! The leach field would probably have sand/gravel in it around the perforated pipes, but if you excavate a bit further, lay the geo-source loop and cover with native soil again before the sand/gravel, it will probably be okay. It all depends upon the local inspectors and what they are willing to allow.
 
@Ampster I'm talking about individual Tesla battery modules salvaged from wrecked vehicles. I didn't realize that LFP was getting that cheap now. I'm assuming you picked up some grade B cells at that price? Even with BMS and other parts that would be cheaper and more usable than my idea with the Tesla modules. Do you have a thread I can read about your battery bank?
They were sold as grade A cells but I am excercising them as if they are Grade B. (80% DOD) I have an Orion BMS left over from an EV project so that drove me toward a single pack with multiple parallel cells. Some people are using less expensive BMSs with multiple parallel packs. That is what would have to be done with Tesla modules but you could put two modules in series with one BMS.
Most of my system design is in my signature but it is evolving. As far as sources for batteries that is evolving also. EVE LF280 were the battery of choice but today I would choose the Lishen 272s. There are many threads here discussing various vendors
 
Here is just one thread about a Lishen build. There are many others:
 
I understand your want to maintain lifestyle while going solar. It WILL be expensive if you are not willing to make any concessions to conservation.

Conservation is most always cheaper than building out solar to support wasteful loads.

No need to go extreme. Go heavy on the insulation. LED lights. Switches or timers for parasitic loads. None of this has a significant effect on lifestyle.
 
For off grid with over 20kwh/day or over 10kw peak i suggest two complete systems.
One for critical house loads (lighting/refrigerator/security etc), and one for high loads (bore pump/reverse cycle aircon etc)

Biggest advantage is you can run the aircon system down to the minimum overnight then shut it down until daylight.
 
In our current 2000sqft home we use about 20-25kW per day. Estimating for the new house, it's looking like 90kW for a high usage day isn't out of the question. I expect something closer to 30-45kW most days. The house will be larger, 6000 sqft with about 1000 sqft unfinished storage, plus a shop with 240v tools. We will be using lots of insulation and paying attention to air sealing, so there shouldn't be a huge amount of cooling and heating load required. Our current home is not highly insulated or really well sealed. It isn't horrible, but it isn't great either. We have natural gas doing most of the heavy lifting today with the furnace, water heater, and stove/oven. Our hope is that the extra insulation and air sealing will significantly reduce the need for heating and cooling along with taking advantage of passive solar heating options.

I am also thinking that prioritizing loads will be in the future if we go for more reliance on PV, but I want to avoid it as much as possible to keep my wife's life simple. Since most days are pretty sunny, and we only use lots of power when we are awake, I'm hoping to oversize the panel array and work on more of a use it or lose it lifestyle. We have plenty of property for as large of a ground mounted array as we like.

Right now I'm looking at the Tesla batteries. I can't get their full capacity with SMA (or any other manufacturer's) gear due to the weird voltage, but they are so cheap I can just buy more to deal with the voltage issue. Though, I just did the battery math on my proposed holiday crazy usage situation and it won't work. I'm planning to start with a 10 Tesla battery module array and assuming I can use 3kW of the 5kW available in each module. So I'd only have 30kW battery available at full capacity which would last for a few hours of insanity (since it won't actually be 24-30kW constant) and then the party's over. I'll absolutely have to plan to have a decent sized generator to handle the loads during excessively high usage and on zero PV days which could correspond with high usage days. And it seems like I'll have to plan a larger battery bank too. Plus think about moving some appliances to propane.

Seems like my dream of limiting propane to a gas cooktop and small generator are dead via math lol.
To crunch a few numbers - by 90kw you actually mean 90kwh which when you break it down to a 8 hour day (even though it'd probably be spread out over a 18 hour day) that gives 11,250 watts continuous so you could use a 12kw inverter to meet all your needs.

12kw means you can have your electric dryer AND your electric stove going at the same time. If you want to run the oven too cooking a turkey dinner you wouldnt be able to run the dryer but I doubt that'd be much of a problem.

So thats the inverter sorted out. The killer will be how much storage.

The rule-of-thumb is you should have 3 days of autonomy of battery power. You can get by if you live in a sunny location with 2 days - I used to live in Clearfield and Layton UT so I know you get a lot of sun both winter and summer, there are the occasional fronts that'll dump snow/rain in the winter but they never seem to last more than a couple days. Still, 3 days is a good target.

You REALLY need to do a good audit of how much power you'll ACTUALLY need.

35kwh/day is a good target, its what I use at home with a gas furnace and water heater and electric everything else and I waste a bunch of power - mainly keeping a 10 year old AGM battery bank charged and floating.

More number crunching to give you an example:

35000wh / 51.2v (LFP actual nominal voltage) = 684AH so for 1 day you'd need 684AH of power with a 16 cell 48v (nominal) lifepo4 bank.
Times 3 days = 2051 AH @ 48V.

To work it out another way, 1 string of 16 280AH EVE LFP cells is 14,336watt-hours or 14.34kwh so you'd need 7.32 strings to meet your 3 day target of 105000watt-hours - round that up to 8 strings and you're looking at 128 280AH cells. At 110 bucks a pop you're looking at $14 thousand in batteries alone.

So now that we've done that exercise you can see that the cost of conservation of electricity is one HECK of a lot cheaper than buying battery power.
 
You can trade off battery sizing to do just 1 day of storage and expect to run the generator more often when peak load is high or solar generation is low. With a battery for handling peak loads, you can actually use a surprisingly small generator.

Assuming a no-solar day, at your current usage of 25 kWh/day and assuming 4-hours a day for the generator, you would only need a 25/4 = 6.25 KW generator which is in the range of portable generators. For your estimated usage for the new home at 90KWh/day, you would need a 22.5 KW generator which is whole-home class generator from the likes of Generac.
 
Heating is always a bugaboo in off-grid living. Unless you run a woodstove - and then you overheat one part of the house and freeze out the rest. A better way is to use a woodstove with a extensive hydronic loop in it that charges a thermal store and the heat is distributed via radiant flooring, hotwater baseboards (what we used in Anchorage, works great with low-temp water) or even euro-style radiators. There are some real downsides to radiant flooring as primary dwelling heat, you'll really want to research that before you go with it.

If you use a phase change thermal store you can pack a LOT of BTUs in a small space and it has the advantage of maintaining a very even temperature until the phase change store has turned completely solid.
You can think of a phase change thermal store as a barrel filled with dozens of copper pipes filled with paraffin wax that the water/glycol circulates around, its a bit more complicated that that but you get the idea.

They even make a water-source heatpump to heat up domestic hot water - it takes the BTUs in lower temp water in the thermal store and concentrates it in the much higher temperature water required for domestic hot water.

As others have mentioned you can charge the thermal store with air or ground source heat pumps using excess panel power. Straight resistive works but heat pumps are a multiple times more productive.

A/C in the summer is easily met by ductless mini-splits - the type that has one central heat pump with multiple individual heads in ceilings of the various rooms, no problem powering it as you'll have power to burn during AC season - and you'll definitely need AC in Utah.
 
Several people do what I am about to suggest, I almost did so but in the end I did not need to.
My Radiant Heating system is a dedicated heat system because I use (Propylene Glycol) in the system and the 7/8" PEX is embedded into my frost-protected slab foundation (FPSF), I considered a few options. I also have a woodstove which I don't use as often as I thought I would have to.

One consideration was to use a Cast Iron Radiator (like those from old buildings that had steam heating) as a collector behind the woodstove which would feed into an insulated water tank that would serve the radiant heating. The tank would also have DC Elements that could take excess generated electricity as a dump load. The Iron radiator would act as a wonderful heat collector and with a reasonable pump it could circulate to the tank and then to the floor from the tank. There is no heat / energy loss in doing so.

My entire Radiant System was supplied by https://www.radiantcompany.com/ who even helped design the layout and everything. All the products & materials used are of the highest quality out there. The preassembled plumbing panel and control systems make it a heap load simpler to install & setup. Their Tech Support is Top Notch and they are always ready to answer any question. Bang per Buck this company cannot be beaten, I spent over a year researching Radiant Systems and Suppliers and these guys win Hands Down.

There are different ways to setup a Radiant System, the two most common are "Standalone/Dedicated" like I have done and with dual purpose of Hot Water & Radiant Heating combined, which let's you use one On-Demand heater or combos of systems.

 
Utah.
6000Sqft house.
Off grid solar
New construction.
I know you are planning to insulate well, but air sealing, and total envelope air exchange is going to be very important for home health and energy conservation.
Will there be a basement? What construction style are you planning? Look ar Matt risenger’s YouTube channel, he is very upscale, but has some great energy conservation designs.
 
Back
Top