Minor Backfeed Issue with Utility

[niktak11]

With the hybrid inverters I'm familiar with, it is always connected to the grid (unless the grid is down or outside the allowable parameters) so it doesn't need to do anything special if there's a large load.

 

[OffGridInTheCity]

Yea, this sort of thing is why I went 100% off-grid from the beginning using UL (mechanically linked relay) ATSs Back feed could only occur if the mechanically-interlocked relays 'somehow' both closed contacts at the same time. So far (6 years) this hasn't happened

In the early days, I experimented with limiter inverters - e.g. that sense load and limit feeding home circuits to avoid back feed like this....



but I was never convinced they were foolproof as much as UL ATSs.

 

[zanydroid]

With the hybrid inverters I'm familiar with, it is always connected to the grid (unless the grid is down or outside the allowable parameters) so it doesn't need to do anything special if there's a large load.

SBU off grid AIO is not supposed to run in parallel (if they designed it correctly)

 

[fmeili1]

An additional sign that these AIO's are in general able to inject into the grid is the error code F60 in the EG4-6500EX documentation, with the description " backfeed protecton".

 

[niktak11]

SBU off grid AIO is not supposed to run in parallel (if they designed it correctly)

Ah I thought you were taking about hybrids. My old LVX6048 behaved like you are describing. I'm not 100% sure what it would do if you overloaded it while grid was available though.

 

[zanydroid]

An additional sign that these AIO's are in general able to inject into the grid is the error code F60 in the EG4-6500EX documentation, with the description " backfeed protecton".

Yup, it’s these “interesting” snippets of documentation that made me assume that a lot of AIO hardware is hybrid in other markets.

 

[zanydroid]

Ah I thought you were taking about hybrids. My old LVX6048 behaved like you are describing. I'm not 100% sure what it would do if you overloaded it while grid was available though.

This conversation is inherently a bit confusing. The off grid AIO in question is doing something bizarre. My theory is that it is some combo of wired it wrong during manufacturing, designed it wrong, or converted it wrong from a hybrid design to off grid AIO

And part of my hypothesis is that most people won’t notice if an off grid AIO was accidentally a GTIL in a trenchcoat that continued to work when disconnected from grid

 

[Zwy]

I'm willing to bet there is no syncing of any kind with these inverters

There is to an extent, the inverters will draw some power from grid all the time if AC input is live to enable a quick switch to bypass, otherwise the switch to bypass would take too long.

I don't believe that power is thru the bypass transfer switch.

 

[RCinFLA]

Grid interactive (hybrid inverter) has an AC input 'pass-through' relay. Some have two AC inputs with two pass-through relays for grid input or generator input. Only one input is used at a time with grid input usually taking priority if it is available,

Inverter is always connected to AC output. Inverter can be active, in which case the AC input pass-though relay must be initially open when AC input is first applied.

In standby mode the inverter remains inactive but the control circuitry pre-syncs to the AC input before closing pass-through relay from AC input.

AC input terminals are initially measured for AC voltage and phasing. With pass-through relay initially open the inverter freq/phasing and AC voltage is slowly adjusted to match phase and voltage of AC input. This can take a couple of seconds, up to 10 seconds, depending on frequency span limits set for inverter.

During sync'g process, the inverter will only slew its phasing at no more than about 0.3 Hz per second rate. This is to avoid disturbing possible loads like AC motors that would create forward or reverse surge currents if AC phase is abruptly shifted.

Due to the limited phase adjustment rate, the wider the inverter allowed AC input frequency locking range, the longer the possible phase lock time and the longer the possible time before pass-through relay closes.

Once inverter syncs to AC input the pass-through relay is closed putting inverter and AC output in parallel with AC input.

UPS mode and battery charging runs inverter active in parallel with AC input. If AC input is connected to grid and grid drops outs, there will be a momentary overload of inverter which it detects and opens pass-through relay releasing from AC input. There is a short voltage glitch on AC output due to momentary inverter overload before pass-through relay is released but most appliances ride through the short glitch of up to 8 milliseconds, or about a half AC input cycle maximum length of time.

Because the inverter is connected in parallel with grid it needs to meet minimum requirements to recognize and release from grid in a short amount of time to prevent back feeding grid under grid failure conditions.

 

[Zwy]

True hybrid doesn't transfer, it syncs then connects the grid into the path. Off grid inverter is going to be either grid sine wave or inverter sine wave

Go look at the diagram, right at the top it states AC is sync'ed before relay is closed.

 

[1201]

To @RCinFLA point this vid shows how a true hybrid works. watch the oscilloscope. From about minute 1 to minute 4. The off grid inverters are incapable of this.

 

[1201]

Go look at the diagram, right at the top it states AC is sync'ed before relay is closed.

Which diagram?

 

[hwy17]

I am now disconnected from the grid. Florida code specifically deals with interconnect but not solar off-grid.

The code in most places covers any permanently installed electric equipment on a home, on grid or off grid.

 

[fmeili1]

Which diagram?

this

 

[RCinFLA]

One of the toughest things for a hybrid grid interactive inverter to do is to recognize when AC input goes open circuit. It can only do this by measuring AC current. If there is no charging or AC output loads on inverter, zero input AC current is a legitimate operating condition that does not warrant opening pass-through relay.

When connected and sync'd to AC input the inverter continually adjusts its phase alignment to stay in sync with AC input. When pass-through relay is closed, inverter is slave to AC input, following AC input voltage and phasing.

If AC input goes open circuit the inverter will gradually wander off frequency because of lack of AC input master phasing control. Usually, the phase locking circuitry is biased to cause inverter to gradually rise in frequency until it goes out of set frequency limit which tells inverter the AC input must have disappeared causing inverter to open pass-through relay. It may take a couple of seconds for this to happen, during which inverter AC output is present on AC input terminals. This is what is happening on scope in video when he opens grid input breaker.

If you remember one thing, if pass-through relay is open the inverter is its own master, if pass-through relay is closed the inverter is slave to follow AC input phase and voltage. Quickly switching AC input source, like between generator and grid, can subject inverter to an abrupt AC input phase shift that can cause high surge current.

 

[1201]

this

Did that come from the manufacturer of this inverter? Because I don't believe that

 

[fmeili1]

Did that come from the manufacturer of this inverter? Because I don't believe that

I don't know but I think it's reverse engineered by someone in the community?! But anyway, because this typical AIO's (at least the ones which are based on these Voltronic Axpert Max models like EG4-6500EX, MPP-LV6548, etc.) are at least "able" to inject into the grid - even if this feature is completely deactivated in the firmware (see the Phocos re-branded model), it's very likely that this schema is somehow ok.

 

[Mattb4]

Sounds like you need minimum load on your meter for it to always read consumption. If they are bitching about 1w "ripple" then hookup 10w incandescent light bulb before the inverter and see if they go away. Or like others suggested set up minimum grid load in your inverter. Sounds like Duke Power is the worst. I backfed my FPL meter for a while and they never complained.

The problem for the OP is his utility company is unlikely to accept any solution that is either completely disconnect or a certified interconnect agreement at this point. He is in the position of having to prove his innocence. I think this is going to be an ever increasing issue for folks wanting to have a DIY solar setup.

That is why I proposed in the past that a true Off grid certification exist for the devices not intended to feed the grid. Ul listed Off Grid label would than be all you had to point to when the Electric company SWAT team showed up to raid your solar farm.

 

[SajgZi]

I don't recall if it was with rtl_433 or rtlamr, but using a SDR and listening to the broadcasts for my own smart meter (ITRON) showed a field that was named something along the lines of BackfeedPower and last I saw it read 44. This was 1-2 years ago. Trying to figure out which it was. I haven't been contacted yet. For my meter, there is indeed a specific field that reads out what is presumably the amount of energy you've backfed to the grid. There are also tamper codes with a variety of outputs. I started rtlamr 2 min ago and I see tamper codes with PHY and values of 00,01,02,03. ENC tamper codes are all 00. I'll update when I find which program reads those signals and what the output is.

Edit: found an example here on the forums - the field is called LastGenerationCount in the 2nd of the 3 message types (the NETIDM type) - https://diysolarforum.com/threads/using-rf-to-read-itron-utility-meters.4038/#post-207584

Edit 2: my LastGeneration is now 55. Obtained by running

rtlamr -msgtype netidm

 

[Banjoman]

Simple fix.. use a chargeverter to kick on when battery soc hits a certain percentage and disconnect all utility from the inverter itself.