First large-scale outdoor powerwall build


New CATL Modules with 20x 3.2v 280ah Prismatic - 17.92kWh - $145-$150/kWh live @ www.batteryhookup.com

harrisonpatm

Member
Joined
Jan 5, 2022
Messages
222
I said in a different post regarding humidity that I would post in a month when I have it assembled, but I realized it's probably beneficial to post now and get feedback on things that I might be able to easily change while it's still disassembled.

I've been lurking and learning on the forum for over a year while I process and gather used 18650s from computer repair stores that give them to me. That was my first goal of the project: I didn't want to spend any money on cells and see what I could accomplish under that constraint. My budget is high on time, low on actual currency. In the past year i've built several small packs for various uses, of various levels of success. I was always waiting to see what kind of size and build I would eventually want for my whole house powerwall. I eventually came across this user's post and decided to take my inspiration from there; I really liked the use of small p packs added a little bit at a time to increase system capacity without having to rebuild the physical storage. This way I could keep collecting my free batteries and add them to the wall when I get decent ones, a little bit at a time, and eventually could have enough for the whole house.

I went with 24p packs, average cell capacity of 2100mah-ish, in 14s strings. By the time I had my design figured out, I had enough to start the build with 2 strings, giving my roughly 4.8kwh of storage right off the bat, depending on the level of discharge I go with.
20221229_141059(1).jpg20221230_153752.jpgIMG_20220820_170441.jpg
The batteries would have to live outside. I eventually got enough salvaged wood from pallet boards and disassembled shoe store shelves to build a sort of mini outdoor shed, with enough space to eventually store 4 strings total. Wood got water- and fireproof treatment, salvaged tin roofing and scrap FRP from my (restaurant) work for extra layers.
20230211_122859.jpg20230211_122902.jpg20230211_131631.jpg20230212_141357(1).jpg

I have holes in the front for ventilation in the summer to dump excess heat. They've got a layer of bug screen over them for debris, 3d printed housings for PC fans go on on half of the openings, and 3d printed regular vents go on the other half for static air movement. I had originally planned to install the fans on the top half, pulling air out of the box and away from the batteries, and the static vents will be on the bottom for return air. Does that sound ok? Or is there another configuration that would work better? I have 8 holes for now, and currently have only 4 fans from my scrap pile of matching size and voltage. If I can get 4 more matches, I would probably add them in, via new 3d printed shrouds, to have 4 fans blowing air in, and 4 blowing air out. Then come wintertime I'll physically plug up and insulate the holes to keep everything in there toasty.
20230219_092416.jpg20230219_094343.jpg20230219_094456.jpg

I left gaps in front and behind the batteries because I currently don't exactly know how to warm the batteries in the winter. Some sort of heating element, some sort of insulation, haven't decided yet. I had unexpected success early making a heating bed for my 3d printer project, and it got me to thinking about some ways that I can DIY the heating for the batteries. I have about 6 inches on all sides of the batteries that I can use later, when I decide what I'm doing. In fact, that's mostly what's putting the project on hold for now: I'm waiting for the weather to warm up enough to let me put the batteries outside without having to heat them. Then I have 8 months to figure out my winter plan for next year.

My house doesn't use that much electricity, around 6-12kwh per day, so if I can eventually get enough cells to fill up the storage, it could have enough for 1-2 days of use, which would be nice. But to start, the primary usage of the powerwall will be backup energy storage for my electric motorcycle conversion. I only have 500w of solar for now on the south side of the house. Last year those were used to charge the bike directly via MPT-7210. It worked fine, but charging was inconsistent, and I could only charge when the bike was plugged in, so the panels were doing nothing if they weren't actively charging the bike. Now with a powerwall setup, and a more efficient true MPPT controller, they will be charging the batteries all the time. I will then use a boost converter when the bike needs a charge, to step up the nominal 48v of the powerwall to the 24s LiFePO4 of my bike.
20230219_094409.jpg

If bike charging works for awhile and I still have excess power that can be used, I plan to wire an output to my house via a 48v to 12v buck converter to directly power a few things in my living room: LED house lights that are already on 12vdc anyway, charging phones and computers, kids toys, internet, ect. I have no plans for a 120v inverter yet; i don't have enough panels or battery capacity to necessitate it. But I gave myself enough physical space in my fuse box, and on my busbars, to add one later if I am able to install more panels.

That's it for now, just waiting on warmer weather to actually install the batteries into their space, they're already balanced and ready to go, everything is wired and fused as much as it can be. Let me know if you have any questions, or if I left anything out. Looking forward to comments! Thanks in advance.

Edit: I realized that I don't have a pic currently of how the series busbars work, but they're directly inspired from the build I liked to in the beginning of the post. Every time I add a string, each pack will be connected in parallel via series busbars, which are already connected to the balance wires on the BMS. So any future capacity additions should work without having to rebuild anything.
 

cak

Member
Joined
Mar 14, 2021
Messages
179
Some context for my comments: I live in an off grid cabin in northern MN(very short sunlight hours in the winter, 2hr give or take) My system is designed around 24vdc. 7s100p packs and I just added my third pack but have been running all winter off two packs so about 10kwh. My system charges off a 3kw solar system and a tiny Honda generator for when the clouds take over. Being off grid everything is designed to be very efficient but I also have a large freezer, cook on induction electric, and have multiple big desktop computers for work from home. Even with all that my daily electric use is 3-5kwh when in conservation mode and on a really heavy use day where the big computers are used all day for two people and we do a baking project I maybe use 8-10kwh.

roughly 4.8kwh of storage
My house doesn't use that much electricity, around 6-12kwh per day
For my situation this last winter 10kwh only gave me one day(two nights/evenings) without sun which was a challenge given how we had a solid month and a half with basically no sun. I ran the generator a lot, about 30 gallons of gas, but every full sun day the battery would be up to the 80% charged point by late morning so I was missing out on potential. On the flip side there are only a few hours of sunlight in the winter so with a bigger battery it can be really hard to fill that last 20% of charge since you enter the constant voltage charging which is slow.

I plan to wire an output to my house via a 48v to 12v buck converter to directly power a few things in my living room: LED house lights that are already on 12vdc anyway, charging phones and computers, kids toys, internet, ect.
I think this is a great way to go about things and there are a couple potential pitfalls to consider. One is how dramatic voltage drop can be over even sub 50' wires when looking at 12-24v. This is why I went with 24vdc rather than 12vdc and even with that higher voltage I calculated noticeable losses for 10A(240w) loads in the back bedrooms only 25 floor feet away. It is pretty easy to find RV type electronics that work with 24vdc(I particularly like the strip lights). If I was building new I would probably run everything AC and invest in more efficient inverters and/or run the raw 48vdc to any DC loads and then have a small buck converter per user. I really like the idea of using the DC directly and when everything is really close like in an RV it does make lots of sense but my mind has been changed when looking at a small and particularly larger home type situations. You have some loss in the inverter but with a good inverter that you would probably be using anyways for the high load devices that loss is likely much less than most wire lengths voltage drop you would have with 12/24vdc.
 

italianuser

Member
Joined
Feb 25, 2020
Messages
640
Wow guys, you surprise me every time.
@harrisonpatm oh I like your plan :love:, congrats! you seem to have everything in the right place. The heating system is a good idea, I'd go for a double system, with a backup heating in case any component of the primary should stop for any reason; could be a manual or automatic backup system but I wouldn't sleep OK without that.
@cak ... You have a full DC system and you measure losses on long cable segments, we had quite a few discussions about voltage losses (with all the formulas), couldn't you fix the problem with larger cables? :unsure: I have a DC system with a 10mt (about 30ft) cable and reduced losses using a bigger cable. It seems a strange idea to go thought an AC conversion. My 5KWh offgrid inverter uses is rated "<55W" in standby mode which is quite a lot, could be as high as 1.32KWh/day without doing anything. And my other small 1000W inverter (ok, it's a chinese one) loses so much during DC-AC conversion. You think we can't save a fully "DC world" in any way?! :LOL:
 

harrisonpatm

Member
Joined
Jan 5, 2022
Messages
222
Thanks for the comments. No plans or desire to bring the house fully off grid yet, so I'm not worried about the battery capacity being up to it yet, at least for now. I was just glancing at inverters today for curiosity, and holy moly 300-1000USD for a midrange model? Definitely not yet. I'll stick with direct DC for now. I could also rebuild or add batteries to the powerwall of a higher quality, to get my cell average up, and I could get more capacity that way too. All a learning process anyway. The fact that I'm using so much free and dirt cheap parts, and all my own labor, makes this a great learning experience over the next year, while I see what my house uses, what kind of solar I can realistically expect, ect.
One is how dramatic voltage drop can be over even sub 50' wires when looking at 12-24v
This is entirely fair. And I won't switch to 24vdc yet, like I mentioned everything in the living room is already at 12. As for voltage drop, one part of my salvage-as-much-as-possible constraints is that I get all my wire and cable from scrapyards. At one point, I got 2ea 500' spools of 12awg copper cable for about 10 bucks... at that price, I can double or triple ply the wire runs in nearly every spot. I can severely reduce voltage drop that way. And at that price, I'm not trying to save money on copper wire anymore. Amazing what great parts you can find in a scrapyard.
couldn't you fix the problem with larger cables?
That's the plan!
 

harrisonpatm

Member
Joined
Jan 5, 2022
Messages
222
The weather started warming up a bit here in Michigan, and I'm getting impatient, I've had these packs balanced and ready to go for a month and I want to play! It might get cold again for a couple more days later, but if it does, I will just turn off PV until it warms up and monitor temps. Definitely still brainstorming on how to warm them in the winter...

20230316_150335.jpg

Hopefully nobody says something like, "wow, looks great!", because I know full well how wonky and messy and untidy it looks! But I can say that everything is safe, fused, connected securely, and I'm wagering that I get bonus points for a dirt-cheap build that's still functional.

Right now I still only have the 500w of solar off 2 panels, so this is still just going to be the backup reservoir for my e-motorcycle. Plans for this year though: I have permits approved for 12 panels on my roof (all that I can fit on the west side, I have no south-facing roof), and I'm planning on what level of connection I want for my house. The DC living room project comes first though.
 

italianuser

Member
Joined
Feb 25, 2020
Messages
640
Hopefully nobody says something like, "wow, looks great!"
Ok, I won't say it look great, I'll rather say congrats man! Good job! :D

For the heating problem in winter I liked those "resistor blankets" (heating beds) because they avoid having to distribute the heat around your large space. With my solutions, instead (inefficient peltier cells or small resistor boxes) I'd have the problem of blowing heat around the box using small fans.
My big concern about the heating would be avoiding the single-point-of-failure, in no case the heating should stop working; that means double heating blankets and double PSU.
 

cak

Member
Joined
Mar 14, 2021
Messages
179
(all that I can fit on the west side, I have no south-facing roof)
East side might be better because then the panels would be in the sun at the cooler time of day which means they will be more efficient. Of course that is assuming your east/west are directly so and that there are not other factors like shade or roof obstructions or weather patterns like regular AM sun.

I think your setup looks beautifully clean and tidy. I don't know how much current I would trust to go through those bus bars though but with your current 500W panels I don't expect it to be a problem.
 

harrisonpatm

Member
Joined
Jan 5, 2022
Messages
222
I don't know how much current I would trust to go through those bus bars though but with your current 500W panels I don't expect it to be a problem.
Can you specify which bus bars? The vertical ones? Those are for the parallel connections, so they dont need to carry much current, just balance.

As for the ones at either end, each is doubled, because they will be carrying full series current.

The bars are 1mm x 30mm, so the cross section is 30mm^2.


Here's one table, there's others, but for example, this one is showing me that 37.2mm^2 is good for 160 amps.

As for the copper on the packs themselves, each pack connects to the next via 12awg copper, 3ea per pack. 12awg is good for 25ish amps, x3 = 75 amps per string, and I have 2 strings going so far, meaning each series connections should be rated for like 125-150amps. The weak point should be the packs themselves: 24 cells per pack, I tested them at 1amp per cell, meaning I shouldn't be pushing it past 40-50ish amps on discharge. Which is why I currently have it fused at 40 amps anyway. All the extra copper on both the individual packs and the busbars (as I have calculated it, anyway) is to combat efficiency losses.

And yet, your point still stands: at only 500w of solar, and only small loads, I doubt I will ever need to pull more than like 30-40 amps on discharge, at least with my loads connected so far. I'm calculating that charging my bike at 1500w is still only 30 amps. In addition, as I load it up with 1-2 more strings, I should be able to either pull more current, or pull the same current distributed along 3-4 stings, instead of the 2 that I have installed currently, stressing my cells less and having them last longer.

But by all means, if my math looks wrong, or I missed some factor on my calculations, tell me, I would love to be able to modify it.
 
  • Like
Reactions: cak

OffGridInTheCity

Active member
Joined
Dec 15, 2018
Messages
2,376
Hopefully nobody says something like, "wow, looks great!", because I know full well how wonky and messy and untidy it looks! But I can say that everything is safe, fused, connected securely, and I'm wagering that I get bonus points for a dirt-cheap build that's still functional.
It's a friendly forum, appreciate the pic! :)
 

not2bme

Member
Joined
Oct 16, 2017
Messages
500
My suggestion is to set up your system for 48V (14S). Right now you're probably buying small and trying out stuff. I wasted on plenty of stuff like that early on too but it was a trial. But I also made the mistake of going 24V first. But back then there wasn't any small 48v mppsolar units, so I bought a 2424LV. Now that unit is no longer in use, because I upgraded to 48v. So there's no upgrade path, and you will eventually end up at 48v. (I still have plans to make a small portable generator out of it but it's one of those projects that will come to fruitation one day...)

Also have my stuff outdoors, in a small shed I built out of crates too... but it is weather protected, with all the charge controllers and inverters in there. By having the electronics, it keeps the shed warm, and more importantly it keeps the humidity out by always being warm. Humidity is a killer as it increases the chance of rusting and that's bad for batteries. Also shed is insulated and the heat generated from inverters keep the batteries warm. But when it reaches 10F or colder outside, there's not enough heat from the devices to keep the batteries at 45F, so I have a small ceramic 100W space heater that kicks in on a temp controller.
 

harrisonpatm

Member
Joined
Jan 5, 2022
Messages
222
My suggestion is to set up your system for 48V (14S). Right now you're probably buying small and trying out stuff. I wasted on plenty of stuff like that early on too but it was a trial. But I also made the mistake of going 24V first. But back then there wasn't any small 48v mppsolar units, so I bought a 2424LV. Now that unit is no longer in use, because I upgraded to 48v. So there's no upgrade path, and you will eventually end up at 48v. (I still have plans to make a small portable generator out of it but it's one of those projects that will come to fruitation one day...)

Also have my stuff outdoors, in a small shed I built out of crates too... but it is weather protected, with all the charge controllers and inverters in there. By having the electronics, it keeps the shed warm, and more importantly it keeps the humidity out by always being warm. Humidity is a killer as it increases the chance of rusting and that's bad for batteries. Also shed is insulated and the heat generated from inverters keep the batteries warm. But when it reaches 10F or colder outside, there's not enough heat from the devices to keep the batteries at 45F, so I have a small ceramic 100W space heater that kicks in on a temp controller.
System is already 14s, 48v nominal.

Heating will be sorted this year, ready for next winter.
 

harrisonpatm

Member
Joined
Jan 5, 2022
Messages
222
I got impatient again, because its still been too cold every day to actually charge the batteries. And it looks like it'll stay cold in my area for the next month. So the night's cold temperatures will stay with my powerwall, which is on the north/shaded side of the house. Bummer.

I had 4 of those small silicone heating pads that battery hookup is selling right now for 6 bucks each. When I got them I didn't realize how small they actually are, bit larger than an iPhone. I thought they would be to small to heat up my relatively large area by themselves, so I just sat on them until this morning's grand idea.
Heating elements.jpg

I had a scrap sheet of aluminum plate that I cut into 4 pieces, Kapton taped the heating pads to them, and wired them with just an on off switch for now, directly to the battery's 48v. Turned them on when the internal temp was about 1 degree C. Came back 20 minutes later, it was reading 20 degrees C. Nice! DIY radiator.

I was kicking myself this morning because just a couple of weeks ago I brought about 20 pounds of random assorted aluminum heatsinks to the scrapyard. Those would have worked a lot more effectively than a plan flat piece of metal. I guess I'll just have to scrap some more!

So that was just a proof of concept to make sure my random supplies would work. Obviously I need to find an electronic thermostat so I can get it to turn on and off at my specified temperatures. Plus, I need to insulate the front and back so it can more efficiently hold on to it's heat without wasting so much energy. I will probably do neither of those things until next winter, because I really only have a couple more weeks of sub-zero temps left, so I think I'll just manually turn it on and off, and wait to build it more permanently next year.
 
Top