NighPV - Batteries to use our solar power at night

elkooo

Member
Joined
Aug 9, 2019
Messages
32
So this is going to be the introduction:

Currently I am still in consideration / battery-harvesting / planning process.

What already exists:
Photovoltaic installation with a 3phase grid tie inverter that runs on around 600vdc (up to 1000 possible) and converts all the solar energy imediately to AC.

Our house has a 3phase 380v system (single phase to neutral is 220~230vac).

The plan for now is to start with a single phase 1kw battery inverter.
I would like to build a 14s32p battery with around 5kWh capacity for starters to gather some expirience and learn how to deal with batteries, bms, chargers and everything else needed to safely operate a powerwall system.
I planto connect a first single phase systemto the one phase that sees the largest nightly power consumption (need to measure all phases for a certain period). The system should then charge at daytime (while excess energy from solar panels is available) and discharge at night through the additional inverter to cover the consumption we have.

Later on, a second battery & inverter combo can be added to a second phase and subsequently the same to the third phase. (Can't use the same battery because of common-ground-problems --> phases will possiblybe connected through the inverters and burn them when operating from the same battery system so better to have separate systems for now... and later on maybe replace the inverter with a 3phase-solar & battery inverter)
I do not plan to touch the PV-Inverter right now since this one has no battery capabilities. If it reaches his end of life, I might replace it with a battery-capable one.

Currently I still test cells & work on the cell holders. I am going to 3d-Print them. Will do 4x4 cells, expandible in one direction in order to do 4x8=32cells/ 4x12=48cells etc...


image_bxbdvi.jpg


Cells will be spot-welded.
I will use nickel strip on one side and glass fuses on the other. Strips and fuses are spot-welded to the cells and probably soldered to a copper busbar made from twisted copper wire strands, as this seems to be the easiest solution.

On a 1kw inverter with a 14s configuration, the battery will see around 20amps peak with gives me less than 1A per cell. I think this should allow them to last for a bit.
 
Its great to have specific goals andwish you success - you can always make new goals as you learn more. For me its been so much FUN to develop power production. I'm heading into my 3rd year and will max out this year as my wife won't allow any more panels -but then I guess it won't be long before I need to do maintenance (excuse to buy new batteries) on my battery bank!

>I would like to build a 14s32p battery with around 5kWh capacity
According to my calculations, your going to need really good cells (e.g. 3000mah+) as 3000mah is 96ah * 48v is (nominally) only 4.6kwh instead of 5.0kwh. I actually use 52v for my calcs - what specs are you basing the 5.0kwh on?
 
OffGridInTheCity said:
>I would like to build a 14s32p battery with around 5kWh capacity
According to my calculations, your going to need really good cells (e.g. 3000mah+) as 3000mah is 96ah * 48v is (nominally) only 4.6kwh instead of 5.0kwh. I actually use 52v for my calcs - what specs are you basing the 5.0kwh on?

Currently the overall average per cell is around 2640mah but I have already around 650 cells tested and have over 460 cells with >2500mah (of which I need 448 for a 14s32p).
So yes, you're right - it will probbably be somewhat below 5kWh (48v x 2.6Ah x 32cells will be more like 4kWh). But I do indeed have alot of >3000mah cells - so if I just use the best cells, it might be closer to 5 than to 4kWh (my excel sheet says, I have 470cells >2500mah with a total of 5'023Wh)... after all this is just theory - I will probably not drain the batteries to their minimum and even set the max. charge voltage somewhat below 4.2v/cell for longer battery life.
My first and primary goal right now is to lower my costs for band-consumption, getting peak energy still from the grid (during nighttime / bad weather).

Our solar system currently doesabout 7kwp and produces around 8500kWh/year but we plan to install 6 more modules with 350wp each giving us a total of ~9kwp by next spring/summer. Should then be enough power to keep us well supplied even during winter times - at least together with a decent battery system :)
 
elkooo said:
Our solar system currently doesabout 7kwp and produces around 8500kWh/year but we plan to install 6 more modules with 350wp each giving us a total of ~9kwp by next spring/summer. Should then be enough power to keep us well supplied even during winter times - at least together with a decent battery system :)
Yes sir, I share your enthusiasm. I'm 100% off-grid and use ATS(s). My PV array is also 7kw (6.885kw) since late 2018and so far in 2019I've hit 9,679kwh PV in and 8,352kwh inverter out/actual-consumption. Mabye 500kwh more this year - depends on clouds! The 14%inverter losses are expected but yie yie yie I'd like to have 100% effecient inverter.

I'm also working to extend my system - more panels, more battery. The house is fully functional (cooktop, hybrid hot water heater, regular appliances, and lots of computer/tv, sockets) BUT I'd like to be able to power a whole house heat-pump. I measured the AC this year and the biggest months were only 1000kwh / month - so another 6kw of PV will get me very close and maybe I can power a heat-pump to actually heat the house a few hours a day even in December. We'll see. In any case, I having a lot of FUN!
 
Generally, sharing batteries between inverters is not a problem, even if they are on different AC phases as the batteries are fully isolated from AC and Ground. Unless of course the manufacturer states otherwise. And best to consult a pro on this.
Certainly not wrong to use separate batteries, but you may(will) have some inefficiencies due to having too much/little energy stored on the wrong phase/battery.
 
ajw22 said:
Generally, sharing batteries between inverters is not a problem, even if they are on different AC phases as the batteries are fully isolated from AC and Ground. Unless of course the manufacturer states otherwise. And best to consult a pro on this.
Certainly not wrong to use separate batteries, but you may(will) have some inefficiencies due to having too much/little energy stored on the wrong phase/battery.

If you have 3 phase net metering on the grid connection (which is very likely for the OP having PV and living in Switzerland) you can just connect the battery to a single phase on-grid inverter to any phase you like and the energy meter should still correctly account for the energy you produced. The 3phase net meter adds up the energy flow for all phases.
 
@OffGridInTheCity:
Using a heat pump is a really ambitious project, since it will draw quite some power! Wish you best of success with this! We do the heating with wood pellets (Switzerland is quite cold over winter time and our central heater does 10kWp with wood... it warms the house and aswell the water for showering etc..). Maybe you could consider having a solar thermal system installed - this is probbably more efficient heat recource with the same amount of surface than a heat pump?

@ajw22:
As I was told, sharing batteries in 3phase systems can burn (especially cheap) inverters, since the phase shift causes much higher voltage loads in the inverters through the battery-connectors. I didn't do any research on this so far but since I would like to have as low costs as possible for starters, I will go with a cheap 1kWp inverter to gain experience and later on update the whole system step by step.
If the inverter is replaced later, I can still use it as stand-alone mobile off-grid system for example (has to be a grid/off-grid capable inverter though).

First prints done on my battery-brackets... see picture attached...


image_cmqomb.jpg
 
@ajw22:
Exactly. I had problems with slightly different cell diameters (mostly due to different shrink wrap I think..) so I decided to add a little adjustability to the cellholders. They are a little tight still but with some minor adjustments they should work pretty much as expected :)
Saw yours too... you seem to have quite a large printer... are they made of PLA?
 
Interesting approach. I made my cell hole large enough to handle thicker shrink wraps. Some cells are a bit loose, but the higher holder (12mm) gives it reasonable stability. Not an issue anyway once fuses are attached.

I use mostly PLA with my "small" CR-10 with 30 x 30 x 50cm print area. I wish I had gotten the big brother with 50 x 50 x 50 !
 
enki said:
ajw22 said:
Generally, sharing batteries between inverters is not a problem, even if they are on different AC phases as the batteries are fully isolated from AC and Ground. Unless of course the manufacturer states otherwise. And best to consult a pro on this.
Certainly not wrong to use separate batteries, but you may(will) have some inefficiencies due to having too much/little energy stored on the wrong phase/battery.

If you have 3 phase net metering on the grid connection (which is very likely for the OP having PV and living in Switzerland) you can just connect the battery to a single phase on-grid inverter to any phase you like and the energy meter should still correctly account for the energy you produced. The 3phase net meter adds up the energy flow for all phases.
enki, this was very helpful!!
I just checked the power bill and phoned with my powergrid provider. Looks like I can run the battery on just one phase as you mentioned and the meter will becalculating if I have excess power or lack of.
So as long as I don't want to go completely off-grid, I can at least on the cost side minimize things even with a single phase system :)
 
So another process update:

Love it, if a plan works out well! The plan:

image_swsexh.jpg


image_urlelr.jpg


Basically the idea was, to have two bus-bars. I use solar bus bar (pre-tinned copper, used to lead the electricityinside a PV module to the connection box). Those are rated for 10A continuous(in a laminated environment). So in free-air use I should be on the good side with 10A aswell.
First I solder short nickel strips to the busbars. I made a simple template for that. Then the strips can be spot welded to the negative side of the cell. Since leaks normally occur on the positive side (seal is there... except you pierce the cell mechanically or it suffers from really heavy corrosion) I want to make sure to have good visibility of the positive cell terminals, this is why the fuses went there. Fuses are also spotwelded.
The welder is made from two old lead-acid car batteries (one for welding, the other one to power the coil of the large 500A golfcart-solenoid and the arduino based timer, helping me to get consistent welding times. Aw, yeah, I made a 16cell pack for starters - just some of the worst cells I had around for testing... around 85Wh all togetherapproximately...
Pictures of all the mentioned details below...


image_ztbeoe.jpg

template to solder nickel strips to busbars


image_lbekci.jpg

negative connectors soldered &ready to weld


image_dauyls.jpg

welding test, fuse ondefective cells


image_jxpcvy.jpg

the welder... :)



image_iaaseq.jpg

positive side soldered & ready to spot weld


image_hmhupv.jpg

negatives welded (was definitely on the hot side... must turn further down next time...)


image_yzhvbe.jpg

fuses welded aswell...


image_kmiwgz.jpg

finished test pack... :)
 
Back
Top