Complete powerwall project review

Stijn

New member
Joined
Jun 29, 2020
Messages
6
As mentioned in my introduction post (https://secondlifestorage.com/showthread.php?tid=9620) I promised a plan of my future (to build) system, and I would like your opinions and what I can improve about it!

I hope you like a long read!

Index:
1) Testing the cells
2) Solar panel arrays
3) Electronics, MPPT, inverters, controller etc.
4) Battery Packs + BMS
5) Outdoor Battery Storage
6) Questions

1) Testing the cells.
Thus far I tested several hundred 18650 cells and have currently 150 cells that have passed al my tests.
- First I do a visual check on rust/oxidation (failed = recycle bin).
- Second I do a voltage check, this must be <1.0V (>1.0V recycle bin).
- Third I do an internal resistance test >250mOhm (<250mOhm recycle bin).
After that they get cued for the test bench.

image_yjgjco.jpg

On my test bench I have 8 pre-charging slots at 1A and 4 adjustable slots (currently 500mA). After charging the cells are discharged-tested on 3 Opus BT-C3100 V2.2 (with fan mod). When the cells are tested they get recharged on 8 slots at 1A. All powered with an old ATX computer power supply and fused in the little box with I/O switch. Cells are regularly checked for temperature and heaters are going to the recycle bin. I also have two 10-Ohm 17W resistor discharge slots in case I have to discharge some cells to 0V safely before recycling them.

Tested and charged cells are stored, <1000mAh, 1000-1800mAh and 1800-2000mAh. Only the<2000mAh cells are numbered and registered in my Excel sheet. Cells with 80%+ capacity of the original capacity still left will be used in the power wall. The lower capacities cells (<2000mAh) and higher capacity cells with less than 80% capacity of the original capacity left are stored and are used for other projects. After a month or so I will remeasure the voltage of the cell with the same multimeter to see if the cell is discharging and how much. Attached a screenshot of my Excel file with all the data.

image_untdqq.jpg


2) Solar panel arrays

Currently there are no solar panels installed but the plan is to start with 9. I dont have a particular solar panel in mind, for my calculations I used an 325Wp solar panel with 85% installation efficiency. During operation 34,2V max, 9,94A max and an open circuit voltage of 41,1V.
These 9 panels are going to be mounted on a ramp (3x3) on the flat roof of the garage, hopefully giving us around 2,5kWatt peak. In the future this will be expanded with another 12 panels also mounted on a ramp (2x6) on the flat terrace roof. The roof on the house doesnt have much space for panels because of a large dormer, only space for 3 panels, maybe 6 with some shadows. All these panels are pointed SW (The Netherlands, northern hemisphere), but I could place another 9 on the NE roof for the morning and midday sun but that wouldnt be very effective and there are some trees there.
So for now the 9 panels on the roof of the garage and in the next couple of years another 12 panels on the roof of the terrace when its build giving us almost 6kW peak total. We are living in a urban part in an terraced house so not much options to place the solar panels unfortunately.


image_fxwjho.jpg

Sketchup drawing above, everything to scale except the roof/dormer.

3) Electronics, MPPT, inverters, controller etc.
I was looking for a complete system with MPPTs and inverters. It didnt have to be a all-in-one system, a modular system also works. Or even preferable actually, if there is a breakdown I only have to repair/replace one part and not the entire all-in-one system.

My eye fell upon the brand Victron, a Dutch company specialized in MPPTs, inverters, battery chargers and all accessories which can communicate with each other. Victron equipment has proven itself in all kind of demanding installations like on trucks, trailers, RVs and boats. It may cost a bit more than other brands or systems but generally I prefer to spend a bit more the first time on good equipment then replacing it a couple of times because of breakdowns or incompatibility.

Below a electrical diagram of the main power distribution of my house and especially everything concerning the MPPTs, inverters, solar panels and other bits like switches, shunt trips, circuit breakers, power cables and communication cables.


image_acatyp.jpg

MPPT: Im going to start with the 9 solar panels, one Victron MPPT (150V/70A) with spare capacity for the other 12 panels. When all 6+12=21 panels are installed I have a theoretical max open voltage of 123,3V and 69,6A @100% just under the max capacity of the Victron MPPT. The future panels on the roof (3 or 6) and maybe on the NE side (9-12) also get an identical Victron MPPT just in case one breaks down I have some redundancy and can swap them depending on which PV set generates the most power.

Inverters: For my inverters I choose the Victron Multiplus-II, it is capable of charging a battery but Im not going to use that function. It does have the capability to communicate with 2 other Multiplus-II units to supply a 3 phase system like I have (and use for our breathing-air compressor to refill scuba tanks) in this house. It also has the capability to use a energy storage system like I plan to use consisting of my 14S100P 18650 battery pack with BMS communicating with the Victron system. When equipped with a current sensor on the main infeed line the system automatically regulates the currents so there is no ingoing/outgoing current from the grid. Im currently installing only one Multiplus-II and will add two later when there are more solar panels installed and the battery pack is expanded to 2P or 3P (14S100P). Im connecting it to the third phase because of the relative constant power draw (500-800W) day and night powering my server rack (UPS/NAS/Server/computers etc.) and other domotics and control systems giving it a extra UPS function when on batteries. On phase 1 and 2 are my ovens (2500W and 3500W) so a lot of peak power there. I picked the 5000W version so hopefully I can be self sufficient one day (5000/230=21,7A) with grid power backup.

Controller: For controlling all the Victron equipment I require a controller, I picked the Venus GX because of the capabilities it has regarding the MPPT/Solar/Inverters and a Energy Storage System. I dont need a direct display or some of the other functions the other controllers have. It also has the capability to communicate with the Batrium BMS system.

Fusebox: The fusebox needs to be expanded with circuit breakers for the ingoing and outgoing current for each Victron Inverter. In case of repairs, defects or other unforeseen times the bypass switch can be used to completely bypass the Victron/battery/solar installation. After the bypass switch the circuit breakers are powering the house and the large circuit breaker for the garage. Every circuit breaker and residual current circuit breaker is equipped with a feedback contact to indicate if a breaker tripped to the PLC and domotica system. And a overvoltage protection of course in case a lightning strike hits the power grid nearby (if it hits the house directly I have more problems then just the electronics).


image_rqubry.jpg


All this will be build in my house below the staircase, the cables from the solar panels from the garage will enter the house (20m cable) and from the roof (10m cable) and the battery outside (8m cable). The cables from the battery (with shunt trip) are connected to the 48V busbar together with the incoming power from the MPPTs and outgoing power to the inverter with knife fuse + holder for each inverter. Furthermore some disconnect switches to disconnect the solar power cables, the Victron controller controlling it all and some wire tray to route all the cables neatly.
The diagonal line across the wall is the line the stairs would be, I removed the stairs for this print screen otherwise you barely would see anything interesting.

image_ygvsxh.jpg

*The wires going into the floor are going into and through the crawl space below the house and a tube dug in the ground going to our garage.
* I didnt draw all the smaller power cables, communication cables and other smaller cables, just the PV and 48V cables.

4) Battery Packs + BMS
In the beginning the system will have to do with a 14S100P battery pack, with a average cell capacity of 2,5Ah I will have a theoretical capacity of 51,8V, 250Ah and 13kWh (rounded). Enough to sustain the day/night load mentioned earlier of 500-800W through the day and night. Later when I gathered more cells I will add multiple parallel packs to this one expanding the capacity to 26kW (2P) and 39kW (3P).

The BMS will probably be a Batrium BMS system with 14 Longmods, ShuntMon. Its not the cheapest out there but I really like the interface and positive stories out there. And although not officially supported it does communicate with the Victron controller according to user experiences online.
I dont think I need the Expansion Board right now, but who knows in the future.

5) Outdoor Battery Storage
In case youve read my introduction, Im also a volunteer firefighter and unlike a lot of people I really dont want to have the 18650 cells operating within the house. The amount of potential energy stored in the cells with the amount of smoke and heat generated when it goes south gives me not much time to escape the house in case of a emergency. Ive been in house fires fighting the fire from the inside but even with a relative small fire in the living room or kitchen the amount of heat and smoke is deadly very fast, even without the cells burning too.
And although the fire will be evenly intense at least the heat and smoke are going up in the atmosphere and are not accumulated inside the house. I rather lose the battery pack and the housing of it along with some plants and garden fencing outdoors instead of all my stuff, house of loved ones sleeping under the same roof. The same for the garage with all our diving equipment and other stuff we have laying around there.

So the batteries will be stored/operated outside of the house. I havent really chosen an enclosure for the batteries yet. I really like the weatherproof cabinets we have in The Netherlands used by telecom, cable, power and fiber providers placed in the streets but they are extremely rare second handed. Buying a new one will cost around a 1000 for a 1000x800x300mm (WxHxD) sized cabinet, a bit to expensive for my taste. I will keep my eyes open for a second handed one but my hopes arent very high.
Ive also been looking at metal garden sheds, they are weather proof and even larger than the outdoor cabinets, but not as sturdy. I dont have the room to place a container, our garden is only 8x10 meters. A small cabinet can be placed in the garden, but a large container is no option. And although Im not a carpenter Im able to build my own box/shed from wood.

The only really big thing thats been bothering me about outdoor storage is the operating temperature. The weather in The Netherlands can vary from -10 degrees Celsius in the winter to +35(40) degrees Celsius in the summer but not really swing more then 15-20 degrees in 24 hours unlike some other places in the world. If I build a cabinet myself from wood I can insulate it and the winters would be more easily manageable with a simple electrical heater keeping the storage above freezing but to install an air conditioning unit is a bit overdone. We dont have an air conditioning in our house so there is no possibility to blow some cold air though a tube into the battery cabinet.

Ive attached a drawing of a insulated cabinet I made with the possibility to store 3P-14S100P battery packs with some room in front of the battery packs for fuses, busbars, switches and pack monitoring. The bottom row is empty in the event some water or other dirt gets in unintentionally. Its just a draft for now and will definitely change until I feel satisfied about it.


image_moznuh.jpg

6) Questions:
Some questions are already in the text above bust mostly not, but to summarize them:
1.1: How can I make sure the cells are the right voltage like 30% charge for storage? My current charges only charges to 100%.
2.1: If my calculations are correct the 9 solar panels will produce a max of 2486Wp (3x3x325W@85%), minus the 700W nominal load the batteries will be charged with 1786Wp. At 51,8V pack voltage (14x3,7V) this will be 34,48A in total and 344mA per cell, without nominal load 479mA. I think this is a great charge current per cell, but when I add the extra 12 solar panels the charge current goes up to 984mA with and 1199mA without the nominal load and this looks a bit high to me. Does this mean I have to extend the battery first before adding the solar panels? Or is it possible to manipulate the charging current for the batteries to an acceptable current and deliver the residual current back to the grid?
3.1: I dont have any specific questions about the third chapter/part for now, just want to know your opinion, improvements and thoughts!
4.1: The cell cutoff voltage varies from brand/cell to brand/cell, some are 2,5V but others 3,0V. Can I combine them as long as I keep the pack voltage above 3V or is it best to exclude them and use a lower 2,7V minimal for example for a little bit more capacity?
4.2: Same question as above but now with the upper voltage. Some cells I tested work great until 4,15V (ish) and then start to warm up with 110mA current still flowing through it never reaching the 4.2/4.3V maximum. Exclude them and go to 4.2V max or keep the max voltage at 4,1V?
4.3: Do I need the Batrium expansion board? I only have the shunt trip and no further equipment so far, am I forgetting something?
5.1: Whats the best temperature to operate the battery packs? During the winter the temperature can be minus 10 degrees Celsius but a simple heater (with thermostat) will help with that if the batteries themselves wont generate enough heat while (dis)charging. But how do I manage condensation? Or the hot days like this week while Im typing this (33-37 degrees Celsius in the shade, 20-25 degrees in the night). Do I need active cooling (like an airco) for the cells or is a simple fan enough blowing in outside air?
5.2: What's the best enclosure to house the batteries outside? Metal shed with wood on the inside on a steel frame, or all metal frame? I'm concerned about the rain getting in, the temperature inside (hot or cold), condensation and although we see them rarely mouse, rats and other local animals.

Only one question left: When I start building, would you like to see regular build updates in a project topic or not?

Ask me anything about the project, materials or choices I've made. Tips, tricks and improvements are always welcome!

Greetings, Stijn.
 
That's a huge write up, nice planning :)
1.1 You could make a discharge circuit with a transistor boosted zener diode (or other voltage reference like TL431) & a big resistor approx 18ohm 10W to discharge cells to approx 3.5V

2.1 You could limit the charge current in the MPPT(s) to flatten out the charging a bit or add more battery capacity.

4.1 Using manufacturers low cut voltage is not the best for cell life & the low end of the curve has little energy anyway. You would be better to set the min at 3.0 or even 3.1V
4.2 Apparently some cells improve after getting cycled a couple of times, then they charge to 4.2 OK. If they still heat after 2 cycles, set aside & use for something else. See the known "heater" issues with the Sanyo red cells.
For "normal" pack use, you would not charge the cells to 4.2 only to 4.0 or maybe 4.1V for better cell life. You should be using the cells in the middle sloping part of their voltage graph, away from the curving ends top or bottom.
4.3 You might want to control heating of cooling direct via the Batrium system (I do that in my system)

5.1 Li-Ion cells like approx 20degC best below 10degC they shouldn't be charged or discharged as hard & below 0degC no charging. At the upper end it's Ok until you hit 40degC but again there's some de-rating from I think about 35degC up.
There are graphs around that show this better than I can explain :)
5.2 Maybe have a look at a small garden shed with a steel frame & polystyrene sheet or other insulation (I did this). The more air tight the better, gaps can be sealed with polyurethane "builders foam". You will need to heat it some when outside is under 0degC maybe 5 degC?

Sure we always like a build story :) You get feedback too!


Here's simple discharger circuit:

image_jbftkz.jpg
 
Admire you're planning.... let me add a battery capacity oriented discussion...

------------------
4) Battery Packs + BMS
In the beginning the system will have to do with a 14S100P battery pack, with a average cell capacity of 2,5Ah I will have a theoretical capacity of 51,8V, 250Ah and 13kWh (rounded). Enough to sustain the day/night load mentioned earlier of 500-800W through the day and night. Later when I gathered more cells I will add multiple parallel packs to this one expanding the capacity to 26kW (2P) and 39kW (3P).
----------------------

I'm 100% off-grid.
Goal1: Consume 100% of PV generated power, which require abattery bank large enough to store/consume the excess PV power generated during the day.
Goal2: Maximize the life of the 18650 battery bank to 20yrs (I know - its a stretch :)) which means targeting a low DOD in the middle voltage range.

I run 12.85kw of PV with an80kwh battery bank. The battery bank has 6 x14s100-120p 18650 lithium-ion batteries at 260ah@48v each =~ 80kwh.
I'm running a 46.3% DOD between 49.5v (3.54v/cell) and 4.0v/cell high with 3.90v/cell average high to be able to consume all the PV power - e.g. the battery provides 33kwh daily on full summer days. 33kwh/daily = 41% of 80kwh and46.3% by DOD. Numbers a bit fuzzy but you get the drift < 50% DOD.

To extrapolate -Take your2.93kw PV vs my12.85kw PV and you get aprox23% PV of my system. If we apply that 23% to my 80kwh battery bank - then you'll need aprox an18.4kwh battery bank for similar results for every 13kwh.

Obviously you're goals may not be the same as mine and your consumption may be more heavily day-time than mine - but still, I think the data I've accumulated is relevant for ball-park thinking.

From the extrapolation - your current battery plans are perfectlyreasonable; however, my comment would be to leave room for a bit bigger battery bank than you're currently planning if you have room. I've found that over time (nearing 3yrs now), and because the battery is the single largest effort (on my part), I'm more focused on making it last! So much work - but its been a lot of FUN as well.

Good luck on building your system.
 
Looks like my outhouse ;)
https://secondlifestorage.com/showthread.php?tid=4229

Batteries are easy to keep warm during the winters. I make do with one small 150W heater for temps under -15C (although mine has the inverters to keep it warm naturally, so you may need to have a larger heater).

The one thing you will need to contend with if you choose to house the batteries separate from the inverters would be the hefty cable length to the batteries. These battery gauge tend to be large due to the current they carry.
 
Looks good Stijn!

Maybe you tossed them already, but if you live nearby Eindhoven maybe we can make a deal?
I get the cells below 1v and you can use my spotwelder? ;) I have a SUNKKO 709A, downside you need a 20A braker for it.
 
Redpacket said:
That's a huge write up, nice planning :)
1.1 You could make a discharge circuit with a transistor boosted zener diode (or other voltage reference like TL431) & a big resistor approx 18ohm 10W to discharge cells to approx 3.5V

2.1 You could limit the charge current in the MPPT(s) to flatten out the charging a bit or add more battery capacity.

4.1 Using manufacturers low cut voltage is not the best for cell life & the low end of the curve has little energy anyway. You would be better to set the min at 3.0 or even 3.1V
4.2 Apparently some cells improve after getting cycled a couple of times, then they charge to 4.2 OK. If they still heat after 2 cycles, set aside & use for something else. See the known "heater" issues with the Sanyo red cells.
For "normal" pack use, you would not charge the cells to 4.2 only to 4.0 or maybe 4.1V for better cell life. You should be using the cells in the middle sloping part of their voltage graph, away from the curving ends top or bottom.
4.3 You might want to control heating of cooling direct via the Batrium system (I do that in my system)

5.1 Li-Ion cells like approx 20degC best below 10degC they shouldn't be charged or discharged as hard & below 0degC no charging. At the upper end it's Ok until you hit 40degC but again there's some de-rating from I think about 35degC up.
There are graphs around that show this better than I can explain :)
5.2 Maybe have a look at a small garden shed with a steel frame & polystyrene sheet or other insulation (I did this). The more air tight the better, gaps can be sealed with polyurethane "builders foam". You will need to heat it some when outside is under 0degC maybe 5 degC?

Sure we always like a build story :) You get feedback too!


Here's simple discharger circuit:

image_jbftkz.jpg



Thank you! I like to plan in advance and hope to not have too many unpleasant surprises when I'm building it.

1.1I really like the discharge circuit, easy to build with aneffective and a consistent end voltage. My first 200 batteries are currently in their 1 month (non)discharge test so when they passed that test I will be discharging them for long-term storage (probably a year or so).


4.1 Didn't really planned on using the entire voltage range, just curious about the possibilities. But it's best to use the mid range indeed like OffGridInTheCity is also mentioning.
4.2 I'm definitelygoing to cycle them a couple of times and see if they still keep heating. The heating usually starts around 4,15V. Would you use them in the pack if the capacity is good and the voltage isn't going to be more then 4,1V or use other cells that aren't heating at all?

5.1 I'm going to check the graphs but with those temperatures and the climate here I'm going to need a heater and a cooler. The outdoor temperature last couple of days and for the next couple of days is around 33-36 degrees celcius (without the extra heat of the cells). Any experience with condensation?

5.2 Thanks for the tip!


OffGridInTheCity said:
[size=small]Admire you're planning.... let me add a battery capacity oriented discussion...[/size]

[size=small]------------------[/size]
[size=small]4) Battery Packs + BMS[/size]
[size=small]In the beginning the system will have to do with a 14S100P battery pack, with a average cell capacity of 2,5Ah I will have a theoretical capacity of 51,8V, 250Ah and 13kWh (rounded). Enough to sustain the day/night load mentioned earlier of 500-800W through the day and night. Later when I gathered more cells I will add multiple parallel packs to this one expanding the capacity to 26kW (2P) and 39kW (3P).[/size]
[size=small]----------------------[/size]

[size=small]I'm 100% off-grid.[/size]
[size=small]Goal1: Consume 100% of PV generated power, which require abattery bank large enough to store/consume the excess PV power generated during the day. [/size]
[size=small]Goal2: Maximize the life of the 18650 battery bank to 20yrs (I know - its a stretch :)) which means targeting a low DOD in the middle voltage range.[/size]

I run 12.85kw of PV with an80kwh battery bank. The battery bank has 6 x14s100-120p 18650 lithium-ion batteries at 260ah@48v each =~ 80kwh.
I'm running a 46.3% DOD between 49.5v (3.54v/cell) and 4.0v/cell high with 3.90v/cell average high to be able to consume all the PV power - e.g. the battery provides 33kwh daily on full summer days. 33kwh/daily = 41% of 80kwh and46.3% by DOD. Numbers a bit fuzzy but you get the drift < 50% DOD.

To extrapolate -Take your2.93kw PV vs my12.85kw PV and you get aprox23% PV of my system. If we apply that 23% to my 80kwh battery bank - then you'll need aprox an18.4kwh battery bank for similar results for every 13kwh.

[size=small]Obviously you're goals may not be the same as mine and your consumption may be more heavily day-time than mine - but still, I think the data I've accumulated is relevant for ball-park thinking.[/size]

From the extrapolation - your current battery plans are perfectlyreasonable; however, my comment would be to leave room for a bit bigger battery bank than you're currently planning if you have room. I've found that over time (nearing 3yrs now), and because the battery is the single largest effort (on my part), I'm more focused on making it last! So much work - but its been a lot of FUN as well.

Good luck on building your system.

I hope toeventually get to a point where I don't use electricity of the grid and only feed the excess back. (Goingoff-grid is not permitted or possible in The Netherlands, you'lll have to be connected to the grid). But I don't have that goal right now, for now I want to store and use all the generated energy without feeding it back to the grid and use is during the night. Because we have a 3 phase 400Vconnection and a couple of large energy users (2 elec.ovens, washing machine, dryer,dishwasher and a 3phase 3kW breathing air compressor) the amount of storage would be huge to never use the grid at all,especially in the winter time. But it remains a long term goal.

But thanks for the comparison, really great the compare the systems (PV and battery size). When I extrapolate to my max size (21 panels, 6,8kW PV) I'll need a ~42,5kW battery. Better start gathering more cells haha.

The storage will be build to house, I think, 3-4 parallel 14S100P packs giving me a theoretical capacity of38kwh or 52kWh which would be right around the 42,5kW I needed according to your calculations.

Thanks!

not2bme said:
Looks like my outhouse ;)
https://secondlifestorage.com/showthread.php?tid=4229

Batteries are easy to keep warm during the winters. I make do with one small 150W heater for temps under -15C (although mine has the inverters to keep it warm naturally, so you may need to have a larger heater).

The one thing you will need to contend with if you choose to house the batteries separate from the inverters would be the hefty cable length to the batteries. These battery gauge tend to be large due to the current they carry.
Looks great! How do you keep the rain out, is it just the plastic or did you cover the roof with something else? I see a litte gap between the door (and inner roof) and the actual roof, is that for ventilation? The rain doesn't reach this part when it's raining horizontally (during storms)?

I do need hefty cables I know, it's a bit of a dilemma choosing between short cables and having the batteries inside the house, or long cables (with some losses) and having the batteries outside. But like I said I really want the batteries outside of the house so I have to take the losses for what they are. That's another reason for going 48V, less hefty cables and less power and voltage losses. Still the cables will need to be like 150-185mm2 (300-350MCM AWG?) but it's only 8 meters so I think the losses are acceptable in this case. But it's going to be a expensive cable (I really hope they have some leftover cables at the company I work at haha).

Bigfoot48 said:
Looks good Stijn!

Maybe you tossed them already, but if you live nearby Eindhoven maybe we can make a deal?
I get the cells below 1v and you can use my spotwelder?;)I have a SUNKKO 709A, downside you need a 20A braker for it.
Bedankt!

I didn't toss them yet, I still have them here (a lot of 0V, or <1V or rusted/leaky ones all in one box) but I didn't count them. But you can have them if you want, otherwise they are going to be recycled (Dutch: gemeentewerf). And you're lucky, it's a short drive, only 20 minutes to Eindhoven from where I live ;-)
I Don't need the spotwelder at the moment (I have a small battery operated one at the moment) but I'll keep it in mind! The 20A breaker isn't a problem, I have a 16AC-characteristic already installed for my work beanch (in stead of the standard B-characteristic), maybe that will do? And otherwise I just swap it out with a 20A or 25A breaker I have lying around. Not really selective anymore with my 25A main fuse but well, when it blows it blows.

Just curiosity, what are you going to do with the cells?
 
>(Goingoff-grid is not permitted or possible in The Netherlands, you'lll have to be connected to the grid).
Interesting. The term "100% off-grid" (in the way I used it)was not ment to say there's no connection to the grid but rather thePV system does not feed power back to the grid - e.g. its 'off-grid' in the sense that the grid doesn't know about the power / there's no feeding it back into the grid. All poweris consumed thru ATSs within the house. The point was to avoid power company permits/complications.

Probably the reason I have the view I do - is that I never considered off-grid/disconnected-from-grid as a reasonable option if there's a grid... because to cover the 3 month winter (in my region) Iwould have to double the panels for those 3 months and the other 9 months they would be too much/useless. So its not practicalwhen grid is available.

Its always interesting to get a wider perspective on terms- thank you:)
 
Stijn said:
Looks great! How do you keep the rain out, is it just the plastic or did you cover the roof with something else? I see a litte gap between the door (and inner roof) and the actual roof, is that for ventilation? The rain doesn't reach this part when it's raining horizontally (during storms)?

For the walls, the outside layer is covered with the blue tarp, it was the cheapest (ie. free) but otherwise I would use a tyvek wrap or something similar. This creates the moisture barrier. Then the wooden cladding goes over it. If any rain or moisture gets past the wooden layers (i.e. between the planks) then it will hit the blue tarp and drain downwards. This creates a way for the structure to dry out and not retain moisture. Behind this tarp will be my rockwool insulation and then another sheet of plywood.

For the roof, the outside layer is covered with blue tarp again or I could use tar paper (or whatever is equivalent nowadays). Then I use shingles for the roofing material, just like any other roof. I found them for free as there was someone in the neighborhood that had some left over and didn't want them.

For the big gap between the door and the roof, yes I suppose I could have done a better job covering it. I meant to put a louvered grill on it but I got lazy and it's been that way ever since. Behind that panel is actually a ceiling where more insulation material is laid on top (for any heat loss above). It also has a round hole in the middle where the fan is used to exhaust heat during the summer or when the temperature hits around 25C.

During the summer the temperature stays about 1-2C above the outside ambient temperature. We only have a few weeks of hot summers that are over 30C so the battery has to work extra hard then, but otherwise it stays within the acceptable levels. It's also located at a location where it's slightly isolated from the wind and also shaded almost all the time.

During the winters the inverters keep it warm most of the time, but there's some -10C times where the heater will kick in to keep the entire structure at 10C.

Hope this helps!
 
2)
The solar panel angle looks a little too shallow. If that's to scale, you'll probably only get near 100% output during noon peak summer, significantly less in spring and autumn, and virtually nothing during winter.
You'll probably want something closer to 45 degrees, but it'll also depend on your local climate (eg. no point bothering with winter, if you get 0 sunshine during winter)
http://www.solarelectricityhandbook.com/solar-angle-calculator.html

Any possibility of attaching panels on the walls? They might do well during winter when the sun is really low.

5)
Absolutely right decision to place the batteries where it can't do much damage. In addition to a heater, make sure there is a sensor inside the box to prevent operation when too cold or too hot.

Q4.1)
The cells will all age differently, some may even suddenly die. So sooner or later some packs will have less capacity than others, and of course the voltage of low capacity packs will drop faster than others. For peace of mind, I'd stop dircharge at 14x ~3.4V, leaving plenty of headroom.

Q4.2)
Lithium cells degrade more quickly when left above 4.1V or so. Most sources seem to recommend max 4.05V~4.1V when high cycle count is required.
I think the rough rule of thumb is that for every 0.1V lower maxV, the usable cycle count doubles (eg. 4.20V -> 500 cycles, 4.1V -> 1000 cycles, 4.0V -> 2000 cycles)

Q5.1)
What must be absolutely avoided is charging at freezing temperatures.
Lithium Ion is super efficient when 4.05V~4.1V max and low current (dis-)charging, so overheating is probably not a big concern, provided the box is not sitting in the sun. My guess is that 2 vents (top & bottom) will probably suffice.

Q5.2)
I have a concrete shed, and keep the doors open during summer. And have a fan with thermostat to provide additional cooling when required.


Definitely regular build updates, including (esp) any failures and lessons learned!
 
Back
Top