External Transfer Switch Bypass Functionality

DIY
1

Hi everyone,

I have a question regarding the bypass functionality of the Victron External Transfer Switch (ETS).

In the Victron documentation, it states:
“No Bypass Switch Required: Eliminates the need for a separate bypass switch.”

Can someone explain how this is supposed to work in practice?

During testing today, I turned off the Victron system, and the ETS contactor automatically lost power on its coil. As a result, the grid was disconnected from the loads.

Am I missing something in the configuration or wiring?
Is there a specific setup required to achieve the “no bypass switch needed” functionality?

Thanks in advance for any clarification.

2

Since that line comes from the external transfer switch manual

i guess your system uses either 8k, 10k or 15k MP2s in parallel. But thats about it when it comes to guessing your system.

What devices do you use? Do you have a schematic?

3

Yes, I’m using 6 units of MP2 15k. The firmware has been updated to version XXX.s99. Everything is working fine, except for that “small” detail :slightly_smiling_face: I used the wiring diagram from the image, exactly as shown there.

image
image1920×1286 611 KB

4

It is designed to inject into the electrical system similarly to AC PV.

You don’t need a bypass as it is not passing through the system like it does in a conventional set up.

What is does need is a contactor that connects and disconnects the other source (i.e. Generator or grid) that opens and closes upstream from the injection into the electrical distribution.

5

So the only issue is that K1 switches off when the L1 master MP is switched off. I guess this is due to the NO contact being used in the Aux relay. I also thought it is a bistable relay internally, but apparently not.

Personally i would add a manual switch that keeps K1 closed, like a Local-Remote switch. Perfect would be to use a switch with two contacts, and use the second contact to create an alarm in the system, so that its activation is not forgotten about.

1 Like
6

Here I’m referring to a bypass in the sense that there is a need to shut down the Victron system, or in case some issue occurs — I don’t want my loads to be left without the ability to be powered from the grid.

Since the contactor coils lose power, it means the grid is disconnected from the loads.

It’s not entirely clear to me why the manual states that a bypass switch is not necessary :slightly_smiling_face:

7

I’m not sure what will happen if the K1 contactor is closed and the Victron turns on its AC-out — will it synchronize with the grid, or will it go boom?

8

Good point actually. ACout 1 cant be switched, its always connected to the inverter internally.

I once supplied AC voltage to the ACout by accident on an MP1, nothing happened, but maybe i was just lucky. I havent tried it (by accident or not) with an MP2. I would assume the devices are safe in that regard but to be honest im not sure.

9

It synchronises from its input only. Device on the output must synchronise to the waveform sent out.

10

Its not. Just the re energised coil for the input is needed, with the connection from the inverters opened to isolate the system.

11

Yes, got it. What’s the best way to implement this?

There would need to be some kind of interlocked mechanism that disconnects Victron AC-out from the busbars while the K1 contactor coil is manually energized — I can’t allow this to be done manually.

What’s the best approach, does anyone have an idea? :slightly_smiling_face:

12

I am designing a system like this (but smaller) and this is also my main concern.

There is one big thing to watch out for when this setup is used; in case of a grid failure, K1 HAS to de-energize. If that doesn’t happen you will be supplying the rest of the street and light up the technicians working to get the problem fixed. So there has to be logic in place that controls the MP-II’s and the override contact for K1.

When you look at the manual for the S99 firmware, it’s mentioned that for some grid codes you’ll need a Ziehl grid relay. I don’t know what part of the globe you’re at, but for a 3-phase setup, I would still recommend using the Ziehl relay. But that is not the issue.

I base my logic around a Finder Opta mini PLC, but a Siemens LOGO will fit just as nicely I guess. This device has 4 relay outputs.

  1. Relay in series with the control of K1 from de MP-II, when this opens The MP-II is unable to close K1 (K1 is 15K1 in my schematic).
  2. Relay parallel with the control of K1, when this closes K1 is closed and can not be opened by Victron .
  3. Relay controlling a small relay switching the measuring voltage to the MP-II’s, simulating a grid outage (20K1).
  4. Relay controlling a contactor for load shedding when the battery state is low (15K2).
  5. The Victrons are in another room where the remote switch is controlled by anaother opta

That comes together in a state diagram like this:

		    5 - LOW SOC (shed low priority)

	 --->	4 - OFF-GRID (not syncing)		         ---
	|							                        |
	|					                        	3z - SYNCING
	|							                        |
	 ---	3 - IGNORE (synced)		    	        <---

	 --->	2 - ASSIST = default (grid conn.)	     ---
	|							                        |
1a - SYNCING					            	1z - DISCONNECTING
	|							                        |
	 ---	1 - SERVICE (grid only)			        <---

		    0 - OFF (all power off)	+ ESTOP state

This diagram will have the following logic;

(f) = forced, (a) = automatic, (Ton) = on delay, (Tof) = off delay

	    0	    1	    1a	    1z	    2	    3	    3z	    4	    5
15K1	-	    x(f)	x(f)	x(f)	x(a)	-(a)	-(a)	-(f)	-(f)    (main grid contactor)

15K2	-	    x(Ton)	x	    x	    x	    x	    x	    x	    -       (low priority circuits for shedding)

20K1	-	    -	    x(Ton)	-(Tof)	x	    x	    x	    -(Tof)	-       (MP-II AC-in (voltage measuring))

MPII	-	    -	    x	    -(Tof)	x	    x	    x	    x	    x.      (MultiPlus remote contact )

So this is my plan, but only in concept. Unfortunately there is no direct solution yet. If you find a glaring hole in my logic, please let me know.

13

Maybe the MP2 OUT will simply synchronize with whichever grid is active at that moment :slightly_smiling_face:

Someone from Victron’s engineering team needs to confirm this for us — or someone will have to sacrifice themselves and test it :grinning_face_with_smiling_eyes::grinning_face_with_smiling_eyes:

14

There was no grid in my example, i simply misread ACin and ACout and fed grid voltage to the ACout.