Disconnecting From Grid

DIY
,
1

We are located in Central Mexico and just had a solar system installed that is both tied to the grid (CFE) and has a battery back-up for use during power outages. The company that did the work installed two Multiplus-II Inverter/Chargers (48V/3000 VA/35 amp), an MPPT 250 i70, a Lynx Distributor 1000 DC and a Pylontech UF5000 battery, along with a 5KW solar array. The issue we are having is that the Multiplus units keep disconnecting from the grid. Yesterday was our first day of full use and it easily did this over 100 times. Whenever I measure the grid voltage I am always seeing close to 120V. Two things seem to trigger it the most - 1)When the solar output goes a bit over 2,000 watts, and 2) when we turn on a higher load such as the microwave oven. It will also do it at other times seemingly at random. My understanding of a system like this is that if everything is working properly the only time the system should disconnect from the grid is when there is a power outage and the battery will be used to supply power for the house loads. My installer (whom I’m quickly losing trust in) insists that this is completely normal behavior for the system. Any advice?

2

100 times disconnecting on a single day is not normal and will wear out AC-in relays very soon. Things that have to be looked at is grid voltage quality and the settings for when to accept grid. Ask your installer for some screenshots of the settings. However it can be really that the grid connection is that bad and in that case another solution has to be looked into like not using AC-in most of the time if possible and rely more on the sun and only maybe help the batteries when grid is more stable during the night? These kind of issues can take some time to figure out the cause.

1 Like
3

Sounds definitely like a bad grid, also noticed you have a really small battery for your system

Recommended minimum is 8kw, if your grid is that bad and you have to go more off grid, you might need even more than 8kw battery

2 Likes
4

The battery is just intended to get us through relatively short power outages. Basically just keeping the lights, fridge and internet working.

5

The battery capacity is not determined by you, but by the equipment manufacturer’s compliance requirements.
However, you are free to act as you wish. Just bear in mind that you may soon encounter other questions.

2 Likes
6

Have a look at the grid voltage and frequency in the Advanced View - it is possible you have supply issues you hadn’t noticed before having the system installed.

The grid-code settings for Mexico might force requirements on your system that are not ideal for your situation.

I don’t know anything specifically about Mexico, but it’s a big country like mine (Australia). It’s possible to be a long way from a transformer and have lots of other customers on that connection - so over-voltage (solar feed-in) and under-voltage issues can be a problem and can trigger disconnects.

7

This is an important reaction
 I have already had several cases where end-users were complaining about the behavior of their system. Too many installers fall for the customer not wanting to pay for the bare minimum needed amount of batteries. Those minimums for a stable system are the bare minimum
 Don’t go under it, don’t sell such a system.

1 Like
8

This situation affirms the “competence” of the installer.
These issues should be addressed by them, selecting the optimal settings for the grid connection.
They were paid for this.

The consumer, however, should reflect on whom they entrusted their energy system to and where they spent their money.

1 Like
9

If the client starts interfering with engineering tasks, they should take full responsibility themselves.
I pass such clients on to others who are willing to take them.

10

For those of you interested, we seem to have solved the issue. We got David Gutierrez (Victron Sales Manager for Mexico and Central America) involved to help our installers. He made some adjustments online to our configuration and the system no longer disconnects from the CFE grid. I am not entirely clear on what he changed, but it is all good now. I do know that we used to see just the sum total of power to/from the grid as well as driving house loads. Now I see it split into L1 and L2.

11

Glad to hear it. I told you straightaway what the issue was and who was at fault. I’m pleased you took my advice on board and contacted the local dealer, who then intervened directly to sort out the problem caused by the installer’s lack of competence.

Good luck.

12

thank you for the UPDATE!!! How can I contact him? we have the exact issue in mexico. Here is the question I was going to post.

System (context)

  • Utility (Mexico / CFE): Bidirectional smart meter (Wasion/Libra3 CHH641). Overhead service.

  • Service type: Appears to be split-phase (two hots + neutral + ground).

  • Solar system Victron: 2× MultiPlus-II 48/5000/120 V stacked for 120/240 V (L1/L2), 2× SmartSolar 250/100 VE.Can, Cerbo GX + Touch. ï»ż ï»żwhen functional for the first 2 months generated 9000 w at max capacity and sent excess to the grid.

  • Battery: 2× Pylontech US5000 (10 kWh), CAN to Cerbo (DVCC enabled).

  • Loads: On solar backup: ï»ż ï»ż ï»żequiment room (400 w) plus cistern cold water pump. other demand, eletricity for house, pool pump (240 V), two wells and one additional cistern pump.

Symptoms (what we observe)

Following the beginning of rainy season . In good sun (~3.5–3.6 kW PV) we often see Grid −3.1 kW (export), then “Grid lost” → export drops to 0 → reconnect → repeats. constant cycling at peak sun.

  • Notifications: VE.Bus error #3 (Expected devices), Grid lost, System time sync error.

  • Voltage indications: Local APC UPS has shown ~136 V at times.

  • Entrance wiring: Field taped splices visible in the outdoor junction can and on the CFE (utility) overhead drop.

1) What the installer recommends

  1. Move the pool pump (240 V) onto the inverter/backup panel to “add load” on the system and (purportedly) improve stability.

  2. Switch to “more stable” lines (change which “two out of three incoming lines”) to feed the inverters to improve performance.

  3. Conductor upgrade has been done already; no change to the taped splices (doesn’t consider taped connectors a problem).

  4. Seasonal explanation: Says it’s the strongest solar time; should “fix itself” late, when the sun is not so strong.

Questions to the community about (1)

  • 1: Does adding a big local load (pool pump) meaningfully help a MultiPlus-II stack stay synchronized to grid or avoid “Grid lost”? Or does it just increase current without addressing the root cause?

  • 2: Is “wait for the season because the sun will not be so strong” an appropriate mitigation for repeated “Grid lost” at 3–5 kW export if voltage rise or input window are the actual triggers?

    3. Is the taped wire connection to the grid “no problem” ï»żor would you find that suspect and worth upgrading?. Victron says this can be a major issue with voltage spkes on the forum, but will not confirm their best practice recommendation.

    4. could the installation of a grid meter be helpful for managing the flow of energy in the system. is this system “hunting” a contributor to the problem?

    5. and if you think what we are asking for is not unreasonable, how do we encourage the installer to change what we suggest as Victron best practices?

    • 6: If we do move the pump to the backed-up panel, what’s the recommended method to prevent battery discharge during outages? (would a 2-pole contactor with a grid-sensed coil, or a Cerbo relay-controlled contactor with rules like “grid present” or “SOC > X%” work?.)

    • 7.: On “pick two of three incoming lines”—if the residence is true split-phase 120/240 (180°), is line selection even relevant on the customer side? Conversely, if a residence is unknowingly on two legs of a 3-phase wye (≈120/208, 120°), what’s the Victron-supported remedy (e.g., Autotransformer + proper config vs re-provisioning utility service)?

      8. would it be helpful to set a little “gate” between 100 W and about 3000 W so that at least some energy can be captured by the system rather than swing and crash all day?

13

Here is David’s email address: dgutierrez@victronenergy.com

14

thank you!! if you find out anything about what he did specifically, it would be appreciated. Also if you have any information on dealing with CFE for any kind of connections improvements, this would be sincerely appreciated!

15

As far as dealing with CFE goes, we have it pretty easy. One of the primary guys in the solar installation company that we are using also happens to work for CFE. I think that has helped things move swiftly and easily with that integration. For instance from the time we applied for our two-way meter, it took just one week for CFE to get the meter swapped. I have heard stories of this typically taking much longer.