Dual PV heat diverted grid backed-up system diagram review !

[Bjorn Teani]

Hi all !

Finally i have my complete diagram. I think I bought the SBMS0 over a year ago…. time flies.

The main objective with this system is to heat 3x100L isolated water tanks and use this water to heat my workshop floor.
And to make use of my panels in the summer I thought I add some batteries to make thing interesting and power some stuff (the office part of my workshop).

The heart of the system will be the SBMS0. And the link to the mortal world would be a Multiplus II 24/3000VA. It’s connected to the grid and thanks to the Virtual Switch it is told to ignore the AC in from the grid as long as :
- the power need are under 2,5KW
- (or) the battery pack is above a certain tension level

I’ll try to be very conservative for the battery settings as I hope this system would last long enough to be profitable. As I’m grid connected (0,165€/kWh), if I’m honest, running on batteries can be a waist of money as I calculated that I need it running as it is for at least 12 years to recover my costs. And that’s without factoring in the cost of the solar panels… But I’m already off topic.

Please review my diagram (inspired from Niels. Thanks Niels !).

The solar panels are SoliTek Standard M.60-B-330 full black monocrystalline 60 cell module

The battery is a 8SP1 3.2v 310Ah Lifepo4.

Here are the part I do not have yet (and I need your input on this) :
- 32A DC Breaker (x12) : TMC 71C 32A
- Battery and PV Shunt : Rideon MKB-300 50mV
- 200A T Fuse : JLLN200.X 125VDC
- 50A Fuse : EATON NON-50

Please feel free to comment on my choices before I hit the buy button.

Bonus questions :
1 - When running on battery, i would be "off grid" for L and N but still tied to the PE of the grid. Is that in some way a security hazard ? I am using differential (30mA) breaker on my "off grid" breaker panel, so I think this would be OK, but I’m still asking.
2 - the Multiplus and the Virtual Switch will be switching the AC in ON i case of a battery voltage below a certain level stopping the drain (it will not be a SBMSO decision…). As I want the grid to kick in seamlessly, I thought the was the best / simplest solution. Correct ?

3 - the Victron BP100 : "Ext IO 3-" goes to "REMOTE +", is that correct ?

Have a great day !

bjorn

[Dacian Todea]

Bjorn,

All looks good except you did not connected the Multiplus to SBMS0 remote ON/OFF controls.

I guess if you disable grid charging then charging will not need to be controlled but if grid goes down then you risk over discharging the battery so at least the inverter ON/OFF control will need to be connected to SBMS0 EXT IOx set as type 2

2 - If you want grid charging then you will need to setup the virtual assistant and control the charging from SBMS0  this will be the document explaining the setup https://www.victronenergy.com/upload/documents/Manual-Connecting-other-lithium-battery-systems-to-Multis-and-Quattros-EN.pdf  all will be as shown there except last to settings will use Disable charger.

3 - Yes it will go to H pin on the remote connector. I think they still use H and L notation.  You can use the same EXT IO3- to go to Multiplus trough another 10K resistor and a 5V zener diode to GND as multiplus remote ON/OFF the Aux i/O are limited to 5V logic and not isolated.

[Bjorn Teani]

Hey Dacian ! Thanks again for taking the time to answer.

So the DEXT connection are all good ? So is the 300A 50mV shunt choice ? The wire gauge ? ... :-)

An you are right, I haven't thought of the case when the Grid goes off. So I will connect EXT IO 6 to the MP. But I will have to find a way to sense if the grid is still on or not so it turns off the inverter only if the grid isn't there. Otherwise I'll want the inverter to keep on powering from the grid. I'll have to dive deeper in the Multiplus manual. Or maybe I just set the disconnect level on the SMBS lower than the virtual switch AC not-ignore it could do the trick : below a certain battery tension, the MP accepts grid input and if its not there and it drain's further the battery, the it will reach the SMBS disconnect level. Yes, I'm thinking out loud.

I although want to thank Oberon for his great work on the Beginner's Guide. Great work !

To all of you reading this, if you see something that could be done smarter (and I'm sure there is) please tell !

[Dacian Todea]

Bjorn,

I have not noticed the wire thickness noted in the diagram but yes 6mm^2 is a good choice. The 300A 50mV is also sufficient for 3000VA inverter and 24V battery.

[Clowny666]

I'm in the process of designing my system and it's always so useful learning from others great designs. I was wondering about the location of the 200a fuse, maybe I am overthinking but if that fuse blows what happens to the DEXT16's? And, should the Multiplus have a separate fuse, if it shorts it takes the 24v loads with it which are probably critical loads?

[Bjorn Teani]

Good point Clowny666 ! Thanks.

So I moved the 200A Fuse closer to the inverter.

I also added the (direct) connexion between EXTIO6 and Remote On/off. EXT IO 6 will be set as type 5 "Fault condition" and should trigger in the situation where I lost grid and battery gets lower than "Low Voltage Chg" set-point (at witch point the SBMS should have already cut the loads attached to the Orion. In normal situations, when getting to the "‘Low Voltage Chg" set-point the Orion should be cut off and the Multiplus should have stopped ignoring AC in.

At least, this is how I picture it... I could be wrong or they could be other solutions. I do not want to use the "Assistant" functionality as I can't use it at the same time as the Virtual Assistant and I do not have found a way to ignore OR not-ignore AC in with it.

[Bjorn Teani]

Hi !

Just a few question as I'm building up my system (still waiting on the battery).

- Is there any way the SBMS0 could monitor the PV Power going into heat diversion ?

- the two resistance on each heating module are paralleled. And they should stay this way, right ?

- I plug a very small battery pack (6s4p 18650cells) and I tell the SBMS I have 8Ah battery pack (to be very conservative), the SMBSO will automaticaly handle the right charge current ? (For now I have the 4 panels from Array 1 and just 1 panel from array 2 connected).

[Dacian Todea]

I have no option at the moment for monitoring the power and energy going in to heat diversion but I will probably offer a external current shunt that will provide SBMS0 feedback about diversion current and voltage either over I2C or using one of the two ADC's

What sort of panels are those ?   You will  need to set a current limit in the DMPPT menu but SBMS0 can not limit the current all it can do is check the current on PV array 1 and 2 and select if array 1 or array 2 or both need to be connected while still being under that minimum set limit.

So say you set a 4A limit but you have a 10A panel on one array and a tow 10A panels on array 2 and if there is full sun the best SBMS0 can do is connect the array with a single panel so you will get 10A even of you set a 4A limit.

Why 6s and not 7s ?  In any case if you have 60 cell panels likely those are around 300W each so you can not connect such a small battery to that system as even that single panel on small array is much higher current output than those batteries can handle. I will say those sort of cells can handle at most 0.25C to 0.3C and so assuming 2Ah cells you will need a minimum of 30Ah battery so at least 7s15p

Not sure how you define array 1 and 2 but array 1 is controlled by the type 1 and should have 1 panel and the array 2 is controlled by type 6 and ideally should have 2 panels not 4 unless you battery can handle 40A for 1 or 2 seconds.

As a resume you should have at least a 30Ah battery then you can have dual array with 1x 10A panel on array 1 and 2x 10A panels on array 2.  

[Bjorn Teani]

The 6s4p that I had laying around is just for testing purposes. And I thought the 18650 could handle 5C (i do not care about longevity).

And I couldn't go lower as 10A for the limit you suggest in DMPPT menu.

I hadn't much of sun the last days so I saw maximum 5A from the two arrays (2x300W + 4x300W). I know that in a sunny day, the small array will already produce around 20A and I was wondering if the SBMS0 would cut both array as even the small one is over the set limit... But according to your answer : no. And I know my battery pack is way undersized but I think the SBMS could "decide" to cut all charge if it can't stay under the set maximum.

I will be looking for your current / power measurement solution on the heating/diverting side, but as you are talking about the "two ADC's", I was wondering if my "old" SBMS0 v.0.2b / 4.0p can follow this evolution as I couldn't find any more ADC's... ?

And I can't seem to distinguish PV1 from PV2 either. Do you confirm that my hardware version is limited in that regard ? Or did I miss something ?

[Dacian Todea]

Is super unlikely any 18650 can handle 5C continues.

There is no lower than 10A limit as it will make no sense since most single panels are already at 10A

The SBMS0 was designed to cut just the large array the small array is the minimum level.

So you will have L1: small array then L2: large array and L3: both small and large. They will only all be off if battery is fully charged.

Yes all versions of SBMS including the older SBMS0 has the ADC's available normally on the WiFi/USB board if you have that option if not you will need to connect to the 16pin connector and add the two divider resistors to access the ADC2 and ADC3

But is more likely that I will want that shunt to be connected over I2C

None of the SBMS0 have anything other than PV1 the dual PV array will not require two separate PV measurements.  The SBMS0 will switch the small array on so L1 and if that drops to half the limit say DMPPT limit is set at 10A then if small array only provides 5A it will test the larger array and if that is below 66% of the same 10A limit so below 6.6A then it will connect both the small and the large array. If then sun will get out and current increases it about a second it will disconnect the small array then if that is not enough the large array and leave the small array only as long as that is above 50% of the set limit. 

[Bjorn Teani]

Thanks again for the time taking answering me (us !). I really appreciate your logic/design/products... but it does present a challenge to navigate the manual(s) and "discussion group" to find all the answers.

And you're (of course) right. I do not remember where I saw 5C... 1C seems to be the maximum. Well, as I said : "testing purposes" and 2C shouldn't do good, but it will not explode (will it ?). I can't wait to get my 310Ah (...) cells to see your BMS in full action !

Waiting also for that shunt over I2C now ! ;-)

[Dacian Todea]

The problem is that cells have large enough internal resistance so that if you try to push 2C in to them voltage at terminals of the cells will increase above 4 or 4.2V limit so charging will just stop immediately. Thus you will think you have a fully charged battery when that is not the case.

[Bjorn Teani]

It's alive !

Well without the LOADS part hooked up...  And some final adjustment (proprely mounting the SBMS0, protecting the battery...)

And now I'm waiting for some sun.

Quick question : Why do I see  LOAD 1.901A when the only thing wired to the LOAD part of the circuit are the DEXTs and DSSRs ?

When I booted up th SBMS I had absolutly no loads (I even disconnected the DEXTs and DSSRs) so I think calibration process went as good as possible. So where do this 1.901A come from ?

[Bjorn Teani]

Too bad I can't edit my previous message. I wasn't fully awake so I had to shunt wires wrong and I forgot shunt Resitance correction...  But now I have another weird voltage reading on cell 1 : it goes up (way more than the other cells) while charging. But a multimeter reading on the cell doesn't show that... I'll search the forum for the answer (and my mistake, probably).

[Dacian Todea]

After you set the shunt resistance please go to Device settings and push the Save device settings button that way the shunt resistance value will be stored permanently even if you power cycle the SBMS0 they will still be there.

Probably cell balancing is enabled and if cell 2 is balanced then during the cell balancing the cell 1 will show a higher voltage due to cell balancing voltage drop.

A photo showing what happens will be helpful.