Rule of thumb for derating Solar Energy off-grid energy systems

[Grinkle]

TL;DR Summary

Looking for a budgetary de-rate to estimate solar capacity needed

I am working on a sanity check for capacity vs load and if anyone here has experience with off-grid solar, any feedback would be much appreciated.

The application is to run a fridge and a modem off-grid on a piece of land in New Mexico. I read that in December (worst case month) the area gets 6.6 hrs of sunshine per day on average.

The stake-in-the ground is to install 600W of solar panels.

End-to-end (panel output to system 12V high-side through a charger) de-rate, intended to include panel over-rating by manufacturer, wiring losses, dirt, charger inefficiencies, etc, I am using 40% loss as my estimate, so I am assuming I will get 360W of usable power for 6.6 hours in December, the worst case month.

This gives me 99W over a 24 hour period, or 2400 Watt-hours per solar cycle.

I will have 600 Ah (@12V) of battery capacity. If I am fully charged on Dec 1, that is 7200 Watt-hours of buffer, less 1800 Watt-hours I need to store each day and use each night (5400 Watt-hours of net buffer) So if I keep the load to 100W average then in the month of December, I have roughly 2.25 solar cycles (2.25 24 hour periods) cumulative of "missing" sun during my expected 6.6 hr charge periods that I can tolerate before the system is drained.

Specific questions from anyone with experience, @anorlunda, perhaps?

Is my thinking on estimating capacity sound? Is my 60% efficiency swag conservative / aggressive / nominal? Of course my need for a fridge in December is not a real thing, but I'm trying to estimate the worst case anyway. All of the other months, on average, should be just fine if December is fine.

 

[berkeman]

Also paging @OmCheeto

 

[anorlunda]

Hours of sunlight are deceiving. When the sun is close to the horizon, the intensity of sunlight is less. Will your panels have a fixed mounting, or will they follow the sun?

What about weather? Thick cloud cover easily overshadows other things such as dust or inefficiency. If your system is really designed for the worst case day, then you must assume thick clouds, or even snow if you are at high altitude.

You said New Mexico, so I assume 30 degrees north latitude. This site can be helpful. For example, this curve of radiation versus time of day, and day of year.
https://www.pveducation.org/pvcdrom/properties-of-sunlight/calculation-of-solar-insolation

What kind of batteries? Lead acid is common and the least expensive, but for long lifetimes, the discharge should not be more than about 40% of the advertised rating. Many people are switching to more expensive lithium batteries today. Lithium batteries can be discharged up to 80% without damage.

Don't forget to include the refrigerator as part of your energy storage system. Make sure that the fridge is as cold as possible just before the sun starts setting. Thermal mass, even water, stored in the fridge helps it to ride it out until the next morning.

But the panels themselves are perhaps the cheapest part of the project. If it was my project, I would just buy 2x or 3x the area of panels that a rough estimate indicates.

 

[Grinkle]

Thanks very much, @anorlunda . That is a great website.

Fixed panels.

Thick cloud cover

I am trying to account for this by assessing how much "missing sun" I can tolerate before going down.

What kind of batteries?

I plan on lead/acid. I need to cut my assessments of reserve power roughly in half - I agree that I want to disconnect the system sooner than fully discharged.

Don't forget to include the refrigerator as part of your energy storage system.

Good point, I'll need to do that. I see there are pretty good sized 12V fridges that claim 50 - 60W running power, so if I can get the duty cycle low that will help a lot. December weather will help with that!

2x or 3x the area of panels

I'll look at over-capacity on the panels and see what that does to the cost. I'd need additional controllers and mounting poles, junction boxes and wiring as well, I haven't looked at that total cost-per-panel including all those things.

Very helpful!

 

[anorlunda]

I'd need additional controllers

No. Just one higher rated controller, and one higher rated smart battery charger. By the way, I might double the number of panels, but not double the battery capacity. The variable is the amount of sun power available.

If you enjoy the challenge of design for the worst case day, more power to you. But if you're looking for practicality, you might accept defeat on the worst of the worst days.

A day with thick clouds, snow, smoke gives you zero solar output. No calculation needed.

A mitigation for an occasional bad day could be the discipline to not open the refrigerator door at all until the sun comes back. That means you can't put food in or take it out, but the food inside is preserved. Think of it as a kind of standby mode.

My friend here in Vermont has this fancy solar array with position tracking plus two Tesla power wall units. I asked him how much energy he gets in December. His answer, "Zero."

Another friend has an old-fashioned engine driven refrigerator on his boat. To cool the fridge, he starts the boat's main engine for 15 minutes. But he only needs to do that twice per day. Sometimes, only once. His secret: a top loading fridge, well insulated. How does that relate to you? He never used battery power to run the fridge at night. His fridge was the entire energy storage system.

Can you measure the kWh use of your fridge? Can you run an experiment to see how much energy the fridge uses if you never open the door from 4PM to 10AM the next morning? Can you add more insulation to your fridge? In another thread, we discussed modifying a chest type freezer to use as a fridge. Would that work for you?

 

[Dullard]

Thermal mass can be your friend, but counting on it can be a problem. In the 'good old days' it was possible to intermittently operate a refrigerator and keep things cold enough. The cooling capacity of a typical refrigerator greatly exceeded the 'steady-state' requirement. Improvements in efficiency have greatly narrowed that gap. Most modern refrigerators don't have the 'extra horsepower' required to make intermittent operation work well - they can't 'catch up' very quickly. Some manufacturers have gone so far as to add a 'turbo' button (variable speed compressor) to allow quicker recovery after warm groceries are loaded while maintaining efficient SS ops.

 

[Grinkle]

zero solar output

Hmmm. I think this argues against extra panels - and it makes sense to me; I think the odds of essentially zero solar for days on end in December are very high. If I really care about never being out of power, maybe an investment in an A/C charger would be the most predictable approach, I have a gas generator that could run the charger if I felt the need to top off the batteries. But I think I am ok with

you might accept defeat on the worst of the worst days.

Regarding worst case, I have given that some more thought.

If you enjoy the challenge of design for the worst case day

I like understanding the worst case, but during this discussion I have come around that I am being a bit silly if I think December is any use case at all for a fridge. It will be outdoors, and anything I put inside it will freeze solid. My realistic use case doesn't start until the average temps are in the 30's at the lowest, and maybe not until they are in the 40's. That leaves me with needing to keep a modem going, and I plan to not use the A/C converter that comes with such devices and wire it directly to 12V, with a DC/DC conversion in the path if that is needed. I haven't looked into that power consumption yet, assuming the fridge will dominate, but I expect somewhere between 10 and 15W, and if it drops out once in a while I'll manage until it comes back up.

Another aspect of realistic worst case is that this is a retreat, not a primary residence, and it gets some winter use, but most of the use is non-winter. Considering December is mostly an exercise in hypothetical understanding, not really an important expected use case.

Regarding the freezer to fridge conversion, if you can link that thread, I'd be interested in reading it.

Can you measure

I don't yet own a fridge, but once I do I can characterize it.

 

[anorlunda]

Can you buy a chest freezer and just adjust the temp to fridge range? Of course one would then need two units.

Yes. It's been done, with reports of successful results. Check out these threads:

https://ecorenovator.org/freezer-fridge-conversion/

https://ecorenovator.org/diy-superefficient-fridge-1-kwh-day/

https://ecorenovator.org/forum/ is a good forum for people who like to think about, discuss, and try these sort of ideas.

Considering December is mostly an exercise in hypothetical understanding, not really an important expected use case.

That probably makes sense. But your questions made me think more about solar plus chest freezer. I'm going to open a new thread on that idea.