Off-grid cabin, large system, starting from scratch

Bluedog225 · November 2019

Hello

I’ve enjoyed learning from this forum for a while and have created an account since I am making progress on my cabin/retirement home and will be installing solar (more below).

Pardon the long post. Short version-I’m starting from scratch and want to design a system that will support full-time living in an off-grid cabin (with robust air conditioning).

I am in central Texas, about 45 miles east of Austin, on a 25 acre wooded lot. Thirty days of 100F-110F days in the summer would not be unusual (lately).

The cabin is 25x25 (625 square feet). Tall ceilings (about 18 feet), with a loft. I have it dried in and am working on installing lots of insulation.

The flat metal roof is double lock, standing seam. 12” o.c. clips on 3/4 ply. About 36x36 (about 900 square feet). Angled slightly down (1:12 slope) to the west/southwest.

Bringing power in will cost about $15,000. Perhaps more importantly, it will require cutting a 30 foot clearing for 2200 feet through the woods at an additional cost. This is not an option for aesthetic reasons. Generator supplementation would be a reluctant option. Wood heat, solar hot water.

Whether this place can be habitable year-round is dependent on air conditioning.

I’ve read many threads on the mini-split systems. I hope to install the necessary panels and battery capacity to make this work.

I’m starting from scratch. Everything is on the table. I’d appreciate some general guidance on the following so I can “do the math.”

1. AC or DC minisplit? I see some energy savings from conditioning without inversion. Is it worth it?

2. What voltage should my system be? 12, 24, 48? Should I be looking at mini-inverters? Do they output 120v AC and allow everything to be wired per residential code?

3. I’ve got the exterior space for any type of batteries. Lead acid would be located about 25-30 feet from the panels. And I can put in a watering system if I go with lead/acid. Lithium could be located much closer. And I’m interested in long-term, robust performance. Lithium seems the way to go. It would be something my wife could continue to use with minimal difficulty if I am no longer around.

4. I plan to mount the panels on unistrut. This will allow 2 season angle adjustment and will give me some space under the panels for air circulation. Is this worth the extra work/detail? Are there better options?

5. What are the “safe” brands of panels, inverters, controllers, etc. By that I mean the Toyota/Honda brands. As opposed to the Lexus/Mercedes or the Walmart brands (sorry to mix metaphors).

Any other input appreciated.

Many thanks

Tom

706jim · November 2019

Panels on the roof or on the ground? If practical, I'd put them on the ground. And if you're planning on A/C, I think 48 volt would be your best bet.

Estragon · November 2019

A big factor in designing a system that will meet your needs is pretty accurately estimating loads. This involves listing all your significant loads, the power (watts) they use, and the run time or duty cycle. For example, a fridge might use 200w running, and have a 1/3 duty cycle, so 200 x 24hr x 1/3 = 1,600 watt-hours/day. A sat-box might use (eg) 30w x 24hrs = 720wh/day. Total all the loads for a wh/day budget.

As the HVAC will be a big part of the budget, you should do the load calcs for proper sizing. Lots of insulation and proper window type and placement etc will be a big help in keeping the size reasonable. Ideally, you will be able to run HVAC to cool the cabin mostly when solar is available and have the well insulated cabin stay cool overnight (rather than running on batteries).

Water and septic can be another big load.

For a full-time cabin you'll want 48v. My understanding is mini-inverters are mainly for grid-tied use, though I think there may be some that can be used off-grid. Generally speaking, I don't see a compelling use for them offgrid.

There are pros and cons to various battery types, but one thing they have in common is not being happy in high heat. If possible, you may want to consider some sort of bermed/buried enclosure for them to moderate very hot temps.

Whether adjustible racking is worthwhile is a personal opinion. I wouldn't roof mount if this is likely (and probably not even if fixed mount). A west-sw slope isn't ideal, and working regularly on a metal roof isn't my idea of fun. You could get a sense of roughly what to expect with various slopes and orientatiins by using a solar calculator for your location, eg: pvwatts.nrel.gov

Pretty much any of the brands sold by our hosts would be considered "Toyota" (or Chev) brands. Don't get caught up in gear yet though, as it's best to get gear that fits in a design. Lots of expensive mistakes get made by fitting the design to gear preferences (or worse, buying gear first).

Welcome to the forum.

Bluedog225 · November 2019

Thanks for the welcome and good info.

I was not aware of any water or septic load. Maybe well and aerated system? Fortunately, we have city water and will have old-fashioned gravity fed septic.

I was worried about the roof direction. Thanks for the link. Very helpful. Unfortunately, the direction of the slope was dictated by other considerations.

Interestingly, the difference between flat mounted panels on the existing roof and turning the azimuth angle to due south is pretty small (13,445 vs.13,332 kwh/year)(0.48 tilt). I’ll need to double check this.

It had never occurred to me to ground mount. Shade aside, I’ve got some security concerns. And all that flat roof 30 feet up with no shade. Although a ground vault for the batteries would really help the heat issue. I’m guessing line losses are lower with 48v? I do have a voided slab and the batteries would be much better off under there. Need to think about this one.

Will work on loads.

Thanks again,

Tom

Estragon · November 2019

Yes, 48v is lower line losses, as,a given power is 1/2 the current at 48v vs 24v, and 1/4 than 12v.

The lower current also means less charge controller capacity for a given wattage array. Eg. A 3.5kw array might get by on a single 80a controller at 48v, but need 4 of them at 12v.

On the AC side, running bigger inverters from 12vdc gets into some unwieldy currents (several/many hundreds of amps DC into an inverter is getting scary IMHO).

mike95490 · November 2019

Roof mount will likely require Rapid Shutdown gear and other complex "safety electronics". The more " safety stuff", the more there is to go wrong.

A pole mount might solve the security issues, that a ground mount has.

48Vm system, no question.

Batteries - as "sexy" as the Li batteries are, they are still not a mature technology, I'd install lead acid, and plan on swaping them out in 5 years with the new lithium systems.

I angled my panels for winter production, in summer I never need to run the generator, but often clouds force generator use in winter.

Insulate the house.

jonr · November 2019

At least take a look at using an air->water heat pump that will allow you to store chilled water in a large tank while the sun is shining. For example, Chiltrix makes one.

Beside lots of insulation, air sealing is critical.

Bluedog225 · November 2019

With the 1:12 slope, it looks like the west south western orientation is not much of a penalty. In fact, it will push my production into the later afternoon, which is ideal. Wish I could claim credit for being smart enough to have planned that.

I like the idea of angling for winter production. Though I imagine I will set up for summer/winter each season. At least till the new wears off.

That Chiltrix looks great. Solves my hot water issue as well. The solar water install was going to be clunky. Thanks!

jonr · November 2019

I wouldn't count on passive battery cooling - I'd either design for AC to keep them cool or put them in a water bath with a small chiller. High temp has a large effect on lead acid life and a moderate effect on lithium life.

mcgivor · November 2019

1. Air conditioning, use a mini split 240V AC inverter type with linear compressor with a high SEER rating
2. 48V would be best, an inverter capable of supporting maximum peak loads with split phase to provide both 120V and 240V
3. LiFePo4 out perform lead acid in virtually every case, they do tolerate higher temperatures but not extreme cold, below 0°C or 32°F They are virtually maintenance free, have a higher initial cost but longer cycle life, can be charged at high current allowing shorter charge periods as well as having no absorption as required by lead acid. They can also tolerate partial discharge state without the penalty of sulfation, perhaps for the non DIY an engineered system with total integration and communication would be the best choice.
4. Roof mounted has its pros and cons, more cons in my opinion, if shading can be avoided with ground or pole mount that would be the preferred choice.
5. Being this would be a serious full time system it would be wise to maintain like components able to communicate with one another including the battery. There are offerings from various sources which are engineered to work together, Outback and Schneider may be the two robust reasonably priced integrated systems available in North America with LiFePo4 partnerships with battery manufacturers, others like SMC, Axicitec , Fronious and Alpha to name a few are not as common in North America, so it's probably best to choose something local.
LiFePo4 by the way is not a technology in its infancy, it has been used extensively outside North America for some years, it's only now becoming realized that it is a superior format there, those with experience will confirm this but are often unwilling to come forward due to push back of those in the jurassic age of lead acid. Sure lead acid has certain advantages, as do ModelT Ford's, but there are few that can't be overcome with available technology. Building a DIY LiFePo4 bank is not that complicated, in many ways it's far less complicated than lead acid with all its quirks and quidditties which often lead to dissapointment. Personally I have both and would never consider lead acid again, it's way too much work especially with a flooded system.

Pardon the weird spacing above, it is more difficult to correct than what it's worth, but reads right despite irregularities.

Bluedog225 · November 2019

Very helpful. Thanks everyone! A couple of more questions.

I’m impressed with the chiller idea instead of a traditional mini-split. It seems more user installation friendly, provides heating (including radiant), and hot water.

Is there a downside I am not aware of?

The battery vault will be outside for lead-acid. Does it also need to be outside if lithium? I’m guessing yes in case of a malfunction and fire.

And lastly, in terms of locating major components. I’m guessing that the higher voltage stuff takes the longer runs? That is, I minimize the runs at 48v dc in favor of longer 120v ac runs. Is that correct? In practical terms, I keep the panels, controller/inverter, and batteries close together.

Thanks

Tom

jonr · November 2019

> Is there a downside I am not aware of?

Cost and the complexity of a hydronic system. Chiltrix isn't a big brand with lots of trained service people.

Estragon · November 2019

The chiller looks interesting. If it gets cold in your location, a glycol fill may be needed. I have glycol in my hydronic heating loops, and it works fine, but the glycol isn't cheap and the heat carrying capacity is a bit lower than plain water. If tanked for storage, you'd want a hx to chill plain water in the tank. Pex is pretty easy to work with. As well as the radiant, I re-did all my potable lines with it.

Panels are usually wired in series "strings", which run at higher than panel voltage, making long-ish (eg low 100s of feet) runs to the array practical. The charge controller(s) buck this down to battery charging voltage. You generally want to keep all the 48vdc stuff (controller, inverter, bank) in one place.

Marc Kurth · November 2019

In a previous life, I designed and installed larger thermal storage systems. (Ice and water) What made them attractive was the ability to use cheap off-peak power rates to run huge equipment to build a cooling supply at night, for use the next day. It worked well in commercial applications because we had all night to run the equipment when the buildings are unoccupied. You will find that they do not make sense for your location/application. (I lived in Bastrop for 10 years)

First of all for cooling, you pay a large efficiency penalty to run a lower evaporator temperature in order to achieve a usable storage temperature. On the days that you need cooling the most, the nights are hot and humid, right through dawn - you have a cooling load 24 hours per day. When I moved there in 2008, we had 96 days in row at 100F and higher. 95F at midnight was not rare. High humidity areas do not have the 30F day/night differential that dry areas do.

It is important to realize that you will need roughly triple sized cooling equipment in order build up storage capacity during solar input hours - to handle the cooling during the hours of low/no solar input. Remember that you are also cooling the home during peal solar hours. Your peak cooling load requirement will be late in the afternoon.

There is much more to it (like maintenance that you specifically said that you don't want) but the basics of operation need to be recognized.

Marc

Bluedog225 · November 2019

Thanks Mark. I was probably understating the cooling requirements. Either I’m in denial or worried folks wouldn't believe me. (chuckle)

It is looking unlikely. 40 panels and 6-10 battle born batteries to run a normal mini-split? Getting too expensive.

Maybe time to look at an underground service from the utility.

Estragon · November 2019

With grid ~1/2mi away, I'd definitely be looking into hookup options anyway.

In any case, I'd also be doing HVAC load calcs. If you're still at a stage where you have options to add insulation, change windows, add overhangs, etc., you can work out the payback periods on the various options.

Marc Kurth · November 2019

How did you come up with 40 modules (panels)?

What do you see as your daily consumption?

Marc

Bluedog225 · November 2019

Rookie analysis follows. Lots of assumptions.

E.g.:

480 ac watts cooling at 24/7 would be sufficient. (I may be able to test this when construction is farther along.)

0.75 is correct “fudge factor” for actual output.

No other loads when cooling.

I’m not a math guy. If there are obvious mistakes, I’d appreciate hearing the input.

If I could get by with a mini split running 24/7 at 480 watts at 120v, I would need 44.16 DC amps x 24 hours=1,060 DC amp hours/day.

30 panels at 315 watts=9,450 watts or 788 DC amps.
(40 panels is the max area on roof. Roof for theft prevention.)

Actual output 9,450 watts x 5.3 avg. hours sun (for San Antonio) x .75 = 37,564 per day dc watt hours.

Or 3,130 DC amps per day.

It looks like I’d be producing excess amps during the day. Enough to run the AC and charge the batteries.

I would need storage for (guessing) 14 hours per day.

14 hours x 44.16=618 amp hours storage needed.

Therefore six 100 amp hour Battle Born 100 ah plus 2 additional for reserve=$8,000 in batteries.

Plus $ for mini split, inverter/controller, install, balance of system.

1500 kwh ac in july. Source: pvwatts.nrel.gov

(not sure how to double check my work using this.)

MarkC · November 2019

Bluedog225 said:

It had never occurred to me to ground mount. Shade aside, I’ve got some security concerns. And all that flat roof 30 feet up with no shade.

Will work on loads.

Thanks again,

Tom

My "solar shed" is in a heavily forested area in my Lavaca Ct., TX, property. I selected a mostly clear area facing South/SouthWest. I am continually surprised how much solar energy is lost due to the surrounding (fairly large) trees especially morning/afternoon and winter conditions. I did not use a microinverter nor DC "optimizer" design, so every "growing branch" can become an issue - including trees that are on the North side of the shed! Careful study of shade (year round) and tree removal requirements would certainly be required in a ground mount design, let alone roof top.

Estragon · November 2019

The basic arithmetic for running a 480w AC load for 24hrs:
480w x 24hrs ÷ 85% inverter efficiency = 13553wh.
At 48v nominal, 13553wh ÷ 48v = 282ah/day. I'd suggest just using watts and watt-hours for these sort of calcs, as converting to amps can be a source of error and doesn't really add much for this.

Very roughly, to replace that, 13.5kw ÷ 5 hrs sun = 2.7kw pv

Whether 13.5kwh/day is the right number though needs to be determined. The load calcs aren't all that hard. Basic numbers just need total square footage of surface area (ceiling, floor, exterior walls, and glazing), temp delta (eg 105°f ambient, 75°f design = 30°f delta), and insulation values for the surface areas. This basic number can then be used to spitball numbers on ROI.

Estragon · November 2019

Optimizers or microinverters aren't likely to help all that much with tree branches. Even bare branches can pretty much kill output if they're close to the panels. The solution is a chainsaw.

I can see where they might be useful working around something like a chimney or vent stack that casts a single shadow that moves with the sun though.

Bluedog225 · November 2019

Thanks. Very helpful. I was making it too complicated.

Are watts the same whether speaking of AC or DC?

My AC load is almost certainly too low. I will work on getting a better number. I’ve been pretty disappointed in my estimations in the past. After I get it closed in and insulated, I may simply put in a window unit and run it for a day off the generator and see what the temps look like.

Estragon · November 2019

Yes, more or less, a watt is a watt. It gets a bit weird with AC power factors sizing loads/inverters etc., but for this purpose it doesn't matter.

Bluedog225 · November 2019

Marc-I reread your post. If I understand correctly, your advice is directed at storage capacity. But a chiller system may still be viable depending on energy use, area to be cooled, and battery storage. Am I understanding correctly?

Marc Kurth · November 2019

Bluedog225 said:

Marc-I reread your post. If I understand correctly, your advice is directed at storage capacity. But a chiller system may still be viable depending on energy use, area to be cooled, and battery storage. Am I understanding correctly?

Yes, Sir - I was pointing out the the pitfalls of storage systems for your application. As you research more, you will find that high efficiency heat pump designs will provide the best balance of efficiency, simplicity and reliability. There have been favorable discussions here about the current breed of high efficiency window units.

I suspect that you will also find that a small, quiet generator will be attractive for occasional use to bridge the most extreme conditions of weather. A generator and propane for heating, cooking and water heating just may keep off grid living viable.

Questions:Is the property in Bluebonnet's territory? What are their rates these days? I am guessing around 12-14 cents per KWH including all charges, for a small user. Have you explored the cost of trenching, instead the 30' wide tree clearing?

Marc

Bluedog225 · November 2019

Thanks Marc.

Yes. It’s Bluebonnet.

I need to look into underground. Last time I checked, the cable alone was surprisingly expensive.

Wholesale Power Charge rate = $0.058936 per kWh

Power Cost Recovery Factor (PCRF) rate = -$0.002365

$56.57
($0.056571 x 1,000 kWh)

+

Bluebonnet Residential Service

The cost of delivering electricity to each member.

Bluebonnet Residential Service = $0.033047 per kWh

$33.05
($0.033047 x 1,000 kWh)

+

Service Availability

The monthly expense for each meter, including billing and payment processing.

Service Availability charge = $22.50 per month

Marc Kurth · November 2019

So, that is $0.09/kwh plus the connection charge. Depending upon your consumption, you will likely be around $0.11 to $0.12/kwh total. That's fairly cheap power. The difficulty you face is the tree clearing eyesore, and only you can put a price tag on that.

Where is your place?

Marc

Bluedog225 · November 2019

Yeah, it is pretty cheap. Will look into digging. They still want 20 feet cleared for buried. I’m off 71 near the park. Have you left the area?

BB. · November 2019

Typically, the "all in price" for off grid solar power is in the $1.00 to $2.00+ per kWH... A few people here, with lots of DIY and managing daily power usage, have gotten down in the $0.50 to $1.00 per kHW range.

Depending on what exactly you purchase (Lead acid, Lithium Ion, cheaper or more expensive batteries, etc.), batteries last something like 5-15 years, electronics typically fail/need replacement after 10+ years.

And if this is a retirement place, at some point if you and your spouse/family can't manage your own system and make repairs/upgrades, you will probably need to hire somebody (if you do not hire one for the initial install).

Also--Look at the value of the property. To me (not in real estate), that spending money on bringing power to the property generally adds to the value of the property. Adding solar to your home, does not seem to add much (if any) value to the property (solar is not, in my humble opinion, an "investment"--It is simply a sunk cost).

For smaller system, Say ~3.3 kWH per day (or ~100 kWH per month), a small system can be cheaper than bringing in utility power (between the initial costs and ongoing connection fees for a weekend/seasonal cabin). For systems larger than ~10 kWH per day (300 kWH per month), these tend to be expensive to install, and may be more than a first time DIY person would want to tackle.

Of course there is the downside of paying the utility fees and monthly power/connection charges. Maintaring the clearing (fire?) for the lines (unless spend extra to underground, or even can underground), etc.

All of this does cost money--But I suspect if you want 12 months a year A/C (heat pump for winter heating?), solar is going to be very expensive (and you have to pay much of the costs for the initial install--At least with the utility, those costs are spread out over decades).

-Bill

Marc Kurth · November 2019

Yes, Sir - we left two years ago. We were off of McAllister Rd.across 71 from the State Park., above the river. Our home made it through the fire, but our dense pine forest didn't. (Same section of woods as the State Park)Then they almost doubled our property taxes to $8,000./yr. The last straw was when they decided that all of our extra land was worth $50,000. per acre, so our taxes were going to go higher. It was time to go!

Now we live on Cedar Creek Lake, Southeast of Dallas.Wish I had done it sooner!

Marc

MarkC · November 2019

Bluedog225 said:

Bringing power in will cost about $15,000. Perhaps more importantly, it will require cutting a 30 foot clearing for 2200 feet through the woods at an additional cost. This is not an option for aesthetic reasons.

Being in Lavaca County (few country miles South), we are serviced by GVEC. I had a similar decision of how to get the power to my cabin - although it is only about 400 feet. GVEC has their own massive "grinder" and was willing to clear a rather winding right of way from the last main pole to my service pole along a path that is now my gravel driveway. Having the 30 foot wide clearing that missed many large oak trees has worked out great and looks really good. Not sure why some providers will do the clearing and include it in their costs, but the San Bernard co-op that is just a few miles away does not either.

Off-grid cabin, large system, starting from scratch

Comments

Categories