diy solar

diy solar

Review my 48V off-grid garage using Victron gear

Mike1000

New Member
Joined
Sep 17, 2021
Messages
12
Hi Guy,

Total noob, here. I was hoping someone knowledgeable could take a look at my plans to set up a solar energy storage system for a detached off-grid garage that I'm building on my property here in Australia. The construction of the new garage hasn't commenced yet, however, I have begun sourcing a lot of the equipment I'll need to build a sweet solar array and battery storage system, while I still have the time. Here are some of the main components that I'll be using or planning to use:

16x 400W Trina Vertex S solar panels - not purchased yet
16x Eve 304ah Lifepo4 Battery Cells (to build a 48v system) - delivered
1x Victron MPPT RS 450|100-TR Charge Controller - delivered
1x Victron Phoenix 48|5000 Inverter - delivered
1x Victron 500A Smart Shunt With Bluetooth - delivered
1x Victron Power In 1000 (with 2x 150A Mega Fuses) delivered
1x JK 16S Active Balancer (2A) and 200A BMS - yet to be delivered (can't wait) :)
... plus various miscellaneous items such as bus bars, fuses, cables, connectors etc.

Anyway, I have attached a diagram below to illustrate how I plan to assemble everything so, if someone could please take a look and let me know what I can do to make the system better/safer etc., that would really be appreciated.

And keep in mind, this system is never intended to be grid-tied. It's purely off-grid. I also plan to charge my EV with it, which is why I have purchased another 16x 304ah cells which are on their way right now (so a total 32kwh battery system is the ultimate goal).

And by-the-way, three months ago, I knew nothing about this stuff. Today, I'm more confident (thanks to all of you and that crazy Aussie, Andy) but I'm definately no expert. Which is why I am here.

I look forward to your comments/feedback.
 

Attachments

  • solar-system-design-garage01.jpg
    solar-system-design-garage01.jpg
    232.8 KB · Views: 134
Victron Lynx power in doesn't support fuses, are you doing this trick?

Nice drawing of a nice design.

You have the shunt and BMS reversed from the orthodox setup.
Its not wrong but worthy of note.
I suggest you take your shunt's positive feed from the system side of the class-t fuse.
That way it will remain powered even when the battery is administratively isolated but won't draw power if something bad happens.

5000 watts isn't a lot for an EV and possibly a electric dryer.
This guy is doing something similar https://www.youtube.com/channel/UCSv0QdWrddSghKQ9gXDT7pQ
 
Victron Lynx power in doesn't support fuses, are you doing this trick?

Nice drawing of a nice design.

You have the shunt and BMS reversed from the orthodox setup.
Its not wrong but worthy of note.
I suggest you take your shunt's positive feed from the system side of the class-t fuse.
That way it will remain powered even when the battery is administratively isolated but won't draw power if something bad happens.

5000 watts isn't a lot for an EV and possibly a electric dryer.
This guy is doing something similar https://www.youtube.com/channel/UCSv0QdWrddSghKQ9gXDT7pQ

Nice drawing of a nice design.
I'm quite proud of that drawing. And thank you for the noticing. :)

Victron Lynx power in doesn't support fuses, are you doing this trick?
Yes, I've watched many Explorist Life videos tutorials (including the one you suggested) which is why I purchased the Power In and not the more expensive Lynx Distributor (which I didn't need). I also purchased all the steel nuts, bolts and washers he recommended, as well as the sizes he recommended.

You have the shunt and BMS reversed from the orthodox setup.
I chose to place the shunt before the BMS because it's the one device that I think most people rely on to give them accurate readings. If I was to place the BMS first, I think I would give me less accurate readings because the BMS would use/draw a small amount of current (for self consumption) that the shunt wouldn't be able to read. What do you think?

I suggest you take your shunt's positive feed from the system side of the class-t fuse.
The reason I placed the small fused wire of the positive wire from the shunt before the T-Class fuse is because it's got its own fuse anyway. But I guess having the positive wire connected after the T-Class fuse is probably better since, in the event of something bad happening, I'll just lose one fuse instead of two.

5000 watts isn't a lot for an EV and possibly a electric dryer.
99% of the charging will be done overnight so I think 5000w should be plenty. The only problem will be trying to get a quick charge. But I already have 240V/32A Grid-Tied charger that I can use for that. So, I think it should be OK.

Andy's Off-Grid Garage YouTube channel has been fantastic for me. Both entertaining and informative... what are great combo!

Thank you for your feedback, SmoothJoey! It was super quick too! Much appreciated.
 
I chose to place the shunt before the BMS because it's the one device that I think most people rely on to give them accurate readings. If I was to place the BMS first, I think I would give me less accurate readings because the BMS would use/draw a small amount of current (for self consumption) that the shunt wouldn't be able to read. What do you think?
If I had a shunt I would probably do it the same order as you.
Thank you for your feedback, SmoothJoey! It was super quick too! Much appreciated.
My pleasure to help.
 
Hey nice drawing and looks like a solid setup. Compare to what I am considering:

Grid assisted off-grid in Costa Rica

I'm no expert on off grid houses but I am pretty good with Victron. Regarding the voltage sense lead from the SmartShunt technically the best place for accuracy is to put it directly on the battery positive post. This lead is non-current carrying and is meant to sense battery voltage. Once a current starts to flow to the Class T fuse there is technically a voltage drop from battery to fuse. Further voltage drop across fuse. That said, if your cable is large enough and your terminations are of high quality and everything is tight/sound it would be negligible if the cable run from battery to Class T is short.

The other thing I noticed is that you are over-paneled a bit which may well be on purpose. Using nominal numbers if you got full output from your PV that would be 6,400 watts. 6,400 watts / 48v = 133 amps which is 33% above the capability of the MPPT controller you speced.

In reality you maybe figure 75% efficiency of 6,400w = 4,800w. 4,800w/48v = 100a so right on the button.

Actually the charging voltage will be more like 56v so that makes the numbers better with 4,800w / 56v = 86a and if you made 6,400w / 56v = 114a.

But 75% efficiency is kind of an average. I'm not sure where you are or what the PV array conditions are but if you have a chance at any point in the day/season to make above 4,800w your MPPT controller may be clipping the peak output and you may be leaving some kwh on the table. Not a problem if this is planned over panelling.

5kw is going to be slow charging for an EV, right? But if overnight I guess speed is not a problem. How many kwh does the EV need? What do you expect for PV production? Making some loose guesses: 4,800w * .75 efficiency * 5hr/day = 18kwh/day expected PV production. If the EV has a 75kwh battery you may need to leave it parked for several days to fully charge the pack and that is if no other loads are using power in the garage. Maybe you plan is just to "take what you can get" not to rely on this to fully charge.
 
Hey nice drawing and looks like a solid setup. Compare to what I am considering:

Grid assisted off-grid in Costa Rica

I'm no expert on off grid houses but I am pretty good with Victron. Regarding the voltage sense lead from the SmartShunt technically the best place for accuracy is to put it directly on the battery positive post. This lead is non-current carrying and is meant to sense battery voltage. Once a current starts to flow to the Class T fuse there is technically a voltage drop from battery to fuse. Further voltage drop across fuse. That said, if your cable is large enough and your terminations are of high quality and everything is tight/sound it would be negligible if the cable run from battery to Class T is short.

The other thing I noticed is that you are over-paneled a bit which may well be on purpose. Using nominal numbers if you got full output from your PV that would be 6,400 watts. 6,400 watts / 48v = 133 amps which is 33% above the capability of the MPPT controller you speced.

In reality you maybe figure 75% efficiency of 6,400w = 4,800w. 4,800w/48v = 100a so right on the button.

Actually the charging voltage will be more like 56v so that makes the numbers better with 4,800w / 56v = 86a and if you made 6,400w / 56v = 114a.

But 75% efficiency is kind of an average. I'm not sure where you are or what the PV array conditions are but if you have a chance at any point in the day/season to make above 4,800w your MPPT controller may be clipping the peak output and you may be leaving some kwh on the table. Not a problem if this is planned over panelling.

5kw is going to be slow charging for an EV, right? But if overnight I guess speed is not a problem. How many kwh does the EV need? What do you expect for PV production? Making some loose guesses: 4,800w * .75 efficiency * 5hr/day = 18kwh/day expected PV production. If the EV has a 75kwh battery you may need to leave it parked for several days to fully charge the pack and that is if no other loads are using power in the garage. Maybe you plan is just to "take what you can get" not to rely on this to fully charge.
Thank you for taking a look at my drawing. Much appreciated.

I have changed the position of the voltage sense, so it's now connected directly to the positive battery terminal and not the T-Class fuse. I have attached an image of how I had connected all the components to make the battery. Unfortunately, the voltage sense cable is not visible, so you won't be able to see the connection.

With regard to the solar panels, I actually ended up getting different panels. They were just too difficult to find so I ended up purchasing 16x 330W Tindo panels (Karra330P). Specs are VOC 40.93, ISC 10.35. I think this should now work better with existing hardware (compared to the Trina panels). What do you think? Spec sheet can be found here: https://www.enf.com.cn/z/pp/98d50a732372a6debbba60b5b61f2ab6d.pdf

FYI, the panels will be installed on a garage roof in Melbourne, Australia.

The EV in question has a 74kwh battery. I travel around 50-60km per day on average. So I think an overnight charge should be plenty.

And since my last post, I decided to purchase another 16x 304 304ah Lifepo4 Battery Cells and build another battery. This will ensure that I have adequate power to run other things as well (small spilt system, external lights, a few tools, etc.). I'll have more details on that down the track.
 

Attachments

  • battery4.jpg
    battery4.jpg
    258.9 KB · Views: 23
  • battery3.jpg
    battery3.jpg
    449.4 KB · Views: 23
  • battery2.jpg
    battery2.jpg
    345.8 KB · Views: 24
  • battery1.jpg
    battery1.jpg
    528.6 KB · Views: 25
Hopefully the voltage sense wire has its own fuse. Just something like 1a to blow to protect the wire if chaffed through or something or if an internal fault in the SmartShunt let the 48v sense get to ground. The wire does not actually have any current except in a fault.

Battery build looks pretty nice to me but I'm no expert in building batteries from cells.

Regarding how it is going to work over time still depends on how much solar production you expect vs. consumption. You now have a little more PV than you originally planned. 16 * 330w * .75 * 5hrs/day = approx 19.8kwh/day production.

You can have all the batteries in the world but if your average daily consumption is greater than average daily charging you will eventually end up with dead batteries each morning and solar will never get them charged.

But if the car only needs a little top up each night and that plus other consumption is less than PV production it should work well. The excess battery will smooth out peaks and valleys in production.

I'm staring to see now why there are so many threads on this forum where someone asks questions and the first response is always "have you done your load/consumption analysis." It is true. How can you build something when you have no idea what the goal is?
 
@Mike1000 I am in the process of dimensioning my so I can't chip in with any input, but thanks for the inspiration, looks awsome! Its very close to what I am looking to implement but with fewer panels.
 
Hopefully the voltage sense wire has its own fuse. Just something like 1a to blow to protect the wire if chaffed through or something or if an internal fault in the SmartShunt let the 48v sense get to ground. The wire does not actually have any current except in a fault.
Sorry for the delay with my response, and thank you for taking the time to comment. The red sense wire actually comes with it own fuse. I'm not sure what rating it is, but whatever it is, I'm sure Victron has added what they think would be the best one to use.

Just a quick update regarding the 2nd group of 16 cells. They arrived a few months back, but I haven't had the time to do anything with it. I've checked all their voltages and internal resistance, but that's about it. Cells are still in their boxes.

Construction of the new garage (it's actually more like a workshop/man cave) is about to commence in the next few weeks. For anyone who might be interested in what it will look like, I've attached a rendered image. Dimensions will be around 6m x 9m (19.5 ft x 29.5ft) or 6.6 squares.
 

Attachments

  • garage-render4.jpg
    garage-render4.jpg
    170.3 KB · Views: 9
  • garage-roof-warehouse-style-mosaic2.jpg
    garage-roof-warehouse-style-mosaic2.jpg
    137 KB · Views: 10
Back
Top