Victron powered house + Man-cave-workshop issue

Q&A and troubleshooting
, , , , , ,
1

Hi All,

slowly making my dream setup. The goal is to power my house and my garage based hobby workshop in off-grid mode for most of the year.

Key information:

  1. my house has 3-phase grid connection (3-N-PE)
  2. house AC loads are distributed between 3 phases but not balanced (not possible to balance, i.e. kitchen 1-phase circuit powers 2kW kettle, 0.8kW toaster, 0.8kW microwave oven, dishwasher (maybe 1kW heater?). Induction cooker is 3-phase but unbalanced.
  3. hobby workshop machines are mostly 3-phase induction motors 400 VAC - the biggest load is air compressor 5.5kW, milling machine 3.7kW, lathe 4.0kW, welding machines (both 1-phase and 3-phased powered). Of course i’m not going to use all of them at the same time but the worst case scenario would be using CNC plasma cutter - this will require air compressor (3ph-5.5kW starting occasionally) + plasma cutter unit (Hypertherm XP45 on 3 phase 400 VAC power on full load consume almost 7kW) + XY table (servo motors + PC, assuming 2kW peaks, couple 500-700 W normal). So this can theoretically draw ca 13kW @ 3ph-400VAC! This will be very rare situation however. What if my wife makes a coffee (2kW kettle) while washing clothes in washing machine?:slight_smile:
  4. i’m going to install 3 air condition units (total max consumption 3-4kW)
  5. i can’t feed energy into the grid - zero feed-in. In fact most of the year i plan to be disconnected from the grid.

Already installed or purchased:

  1. Three Victron Multiplus II 48/5000-70-100 units
  2. 6 x 420Wp PV panels (facing West) connected to Victron MPPT 150/35 charger
  3. 10 x 500Wp PV panels (facing East0 connected to Victron MPPT 250/85 charger
  4. experimental DIY battery (ca 5-7kWh usable energy, old Chevy Volt Li-Ion 12S modules)
  5. Victron Smart Shunt 500A/50mV installed on the battery
  6. Cerbo GX + 7” Touchscreen
  7. 32 x 324Ah LiFePO4 cells waiting for DIY enclosure
  8. 4 x JKBMS 300A V19 and other bits to make four 16 cells banks

Plan to buy/install:

  1. 32 x 324Ah LiFePO4 cells for total 64 pieces system (ca 66 kWh of energy total, in 4 banks)
  2. install additional 500Wp PV panels - 12 or 15 pieces on East side and 12 or 15 pieces on West side roof (35 deg inclination)

And here comes the story - to add those additional PV panels i can see two scenarions:

A) DC coupling by adding MPPT SmartSolar charger/chargers

B) AC coupling on AC-OUT side of the MP II’s using Fronius Symo 12.5-3-M

From what i’ve read here and there, for option A pros are:

+ more efficient charging of the battery bank (DC-DC)

+ not loading MP II because MPPT chargers load energy to battery directly

+ system doesn’t care about unbalanced loads on AC

Cons are:

- AC loads are powered by MP II’s only (grid disconnected/off-grid scenario) so MP II 48/5000 can deliver “only” 4000 W / 5000 VA per phase (i’m starting to regret i didn’t buy MP II 48/8000 units…)

For scenario B pros:

+ much more load available in sunny days as i can theoretically get 15kVA from MP II’s + 12.5kVA from Fronius

+ driving AC load using solar power directly from Fronius is more efficient

+ i can but second hand Fronius quite cheaply

And now how about B scenario cons?

- grid connection preferred to be able to dump excess energy spikes (when battery is full, sunny day and big AC load suddenly stopped)

- frequency shifting (Victron-Fronius “communication”) is slow to handle situation described above

The biggest concern is AC load balance - lack of balance to be precise. I heard that AC coupled Fronius is not an option in off-grid scenario because the way Fronius works - it needs balanced 3 phase load. Is it truth? My case will never have all phases balanced.

Some notes:

  • no, i’m not using air compressor and plasma cutter every day. In fact this is rare situation
  • priority would be to have reliable power for house loads. 99% of the time in the year my garage workshop is not in use

Final thoughts i have now:

  • is it better to consider connecting grid on MP II’s AC-IN’s to allow Multiplus to pass the energy from the grid in case i use big load? I’m afraid Victron will feed-in some energy into the grid occasionally due to the way it is designed (even if zero feed in is set) - is it truth?
  • to balance 3 phase loads i can use air condition units (one unit per phase) + some Nore-RED logic to switch them individually on/off as required to help balance the load

I’m really confused regarding limitations of using AC-OUT side AC coupled Fronius in Victron micro grid system.

Thank you for pointing out weak points in my assumptions, guys!

2

Lots of questions there.
Might be better to split questions up a bit.
Think you might of over estimated the power inverters can supply, 5kva can supply just over 4kw and thats without any thermal derating, yes they can supply more but only for a short time.
Running at close to max will require good cooling of equipment.
Being grid tied in this case would stop any over loading( surge from equipment start up comes to mind) with any deficits of power from inverters being supplied from grid.
Keep us all informed as system progresses.

3

You need a bigger battery :wink:

4

The frame i’m designing will be capable to accomodate 6 pack 16 cells each - i’m using 324 Ah cells, this will give 99 kWh of total energy, 80kWh usable. For the moment my experimental installation looks silly:D:D

experiment2
experiment21417×869 201 KB

experiment1
experiment11920×864 212 KB

Interesting thig is i’m running the cells without BMS(!) - checking cells voltage drift from time to time. Battery is 5 modules, 12S each. Cells are old and have been used when i got them (measured capacity 33Ah vs factory 45Ah). Max battery voltage set in SmartSolar chargers is 12*4.1 V = 49.2 V. I have just measured the cells after one year of operation:

dV after one year without BMS
dV after one year without BMS1760×1553 474 KB

For me pretty amazing results. Yes, i know it’s not the smartest way to run Li-Ions without BMS. But it looks like they are pretty stable.

5

We build our batteries exclusively from Porsche Taycan modules (6S LG_Chem 22V 128Ah 2.8kWh nominal), 2 modules in series per string and upto 16 strings in parallel. Although this is on lower end of the voltage range for the 48V nominal MultiPlus II system and requires special care with the parallel wiring and fusing, it works like a charm. The biggest trick is to make sure all modules are very well matched and balanced to begin with, but then we hardly ever see any cell imbalances develop even over prolonged periods of dayly cycling more then half of the full batterybank capacity (in a DESS Trade only system). And on the rare occasion we do see an imbalance (yes we do monitor for it, but we do not balance automatically) we usually rebalance that string once, only takes a day normally. If the same cells/string develop an imbalance again we change out that particular string and mark it as B grade for other use cases. But normally in a well matched batterybank we don’t expect to see any imbalances develop over long periods of time, a year or longer even.

6

It basically matches my observations.

7

You still do need a BMS though, even if not for balancing, any battery should have it’s own dedicated protection system that disconnects the battery in under/over voltage, current and temperate conditions. As an absolute minimum quality fuses (high ‘breakable’ current) and a reliable voltage/current monitor driving a solenoid/relay.

8

Already have 4 pieces of JK BMS v19 rated for 300 Amps. Final battery bank will have all the bits required. Ceramic fuses are on my desk as well as 250 A rated 2-polke MCCBs.

9

Sounds good. Do lower any expectations you may have on the ability of those JK’s to do reliable SoC calculations. You save yourself heaps of problems using an external battery monitor for that.

10

Already have Victron smart shunt (500 A version) - i plan to use this guy for SOC. Any good idea?

11

Check those MCB’s not only for the rated current but also for the ‘breakable’ current. If possible try to calculate the internal resistance of the battery protected by the MCB, then assume an external short circuit (who has never experienced creating an unforeseen ‘wrench’ wiring circuit) to calculate peak current. This can be 1.000’s or 10.000’s of Amps with ease. The MCB need to be able to break that current without simply arching the current through. If the MCB cannot do that, it needs to be protected against that failure mode with another fuse that can.

12

Yes. I have read before but cannot recall directly, a discussion on how to combine JK BMS data with SoC from the BMV. You’ll have to dig into that yourself. There seemed to be an issue having to choose either BMV or JK BMS as battery monitor at system level, not sure where things stand on that topic.

13

the MCCB i have in hand has ultimate short-circuit breaking capacity @ 50kA and service short-circuit breaking capacity of 35kA. Disconnecting both negative and positive poles simultanously - should be fine. In addition i have big ceramic fuse in series.

14

Care to share brand/type/cost/supplier?

15

Sure! It’s not the fancy one, however i’ve seen some tests and internal built - looks promising. TOM7Z-250/2300. It’s pretty big and heavy. It was ca USD 45 if my memory serves me right.

16

I have 3x MP2 5K 3p and 56kWh battery.

8,5kWp PV are distributed to 4 panels 425Wp on a MPPT 150/45 and 16 panels on Hoymiles micro grid inverters on ACout.

Frequency shifting works great. On grid disconnect/failure if battery is 100% SoC, frequency is increased and the PV inverters stop. Than frequency is lowered so the AC inverter produce “save” below current loads. The fine tuning is done by the MPPT. Its power is adjusted fast, while the PV inverters are mostly constant.

I would go for the Fronius if you can have long strings, else for micro inverters with one MPPT for each panel, max for two panels like the Hoymiles 3p with 3 MPPTs for 6 panels.

Integration can be done with OpenDTU.

17

Appreciate you input, Bjorn. I was considering microinverters but i was not sure if there microinverters ready to be integrated with Victron system. My biggest concern i guess is load balance between phases and ability of AC coupled inverters to react to this imbalance and sudden load changes on AC side. My first assumption was zero feed-in to grid (due to various reasons i can’t register my system to national grid). For this i want to keep MP2’s disconnected at AC-IN side. But in case i will end up having DC coupled system only (all PV panels connected vis Victron SmartSolar MPPTs to power bank) the bottleneck would be MP2’s - unless i use AC-IN to deliver more juice on power surges. In the second scenario - ending up having AC/DC coupled system (some PV panels via MPPT (already have them) and some PV panels via, say, 3-phase Fronius on AC1-OUT of MP2’s) the concern is Fronius ability to provide power to unbalanced loads - imagine the example my house requirements are L1 = 500W, L2 = 2500W, L3 = 7500W while 100% battery in sunny day, disconnected from the grid. L1 and L2 can easily be services by L1 MP2 and L2 MP2, but L3 can only be feed ca 4kW by L3 MP2 and the rest should come from Fronius inverter - but only on L3! Can Fronius do that?

What about using 1 phase microgrid inverters and divide PV panels into 3 sections for the phases? The bottleneck would be available power per each phase will come from 1/3 of PV panels.

What i’m looking for is the ability to redirect all the available solar power from AC coupled inverter/inverters to either L1, L2 or L3 if suddenly required.

I hope i explained my thoughts in the correct way so you guys can get my point.

18

@Bogi

I don’t think the Fronius can unbalance as it has no information about the loads.

I have mixed single phase micro inverters unequally distributed to three phases.

L1: 1x HMS-2000, 1x HMS-1000

L2: 1x HMS-2000

L3: 1x HMS-2000, 1x HM-600

My loads are also unequal and not necessarily matching the inverters. The MP2 have to equalize.

Yes, you need to match loads and inverters.

If you charge you EV with “surplus” PV from AC-PV this would relieve the MP2s. Same applies to other manageable loads.

Hope this helps a little.

Micro grid inverters are not read integrated. There are modifications on GitHub doing the integration. With Venus OS 3.6x SunSpec has been released. I expect more inverters to be integrated. I request SunSpec for OpenDTU but there is no progress. Virtual devices with MQTT might be another option. With Victron there are many possibilities for DIY.

19

Are micro grid inverters fast to adapt to the grid? Could they be switched over to the highly loaded phase fast enough to prevent an actual overload condition?

20

@UpCycleElectric

You have to obey factor 1 rule per phase, e.g. for a MP2 5K not more than 4000W micro grid inverter.

You cannot switch them phases.

The react on frequency shifting when off grid or grid disconnected.

Currently there is no option to throttle them if there specific data transfer unit does not support SunSpec.