How to AC couple Enphase IQ8 w/Battery Storage to Schneider XW Pro 6848 w/rack batteries, off-grid?

Not SB, SBS Sunny Boy Storage. SMA appears to have done the hacking, looks like a Sunny Boy but accepts high voltage lithium battery (really low voltage plus bidirectional boost converter).

It can be in some sort of network with Wattnode, participating in zero export and peak shaving. I think -SA means it will throttle down its output power if line frequency increases. (It also is grid-forming, can do frequency shift to control Sunny Boy.) But beyond -SA, maybe it only responds to Modbus, CT, and time of day clock.

I haven't read of it being used in a Sunny Island system. I think ramping down watts with increasing frequency is already there. What I don't know is if it would continue by ramping up charging in response to further increasing frequency. I would want both of those to occur before reaching the frequency where Sunny Boy curtails PV production.

Max Inrush Current Scenario
- no load on backup panel and max PV backfed into grid
- grid fault
- inverter/charger needs to be able to divert the AC coupled PV energy into the battery, battery cannot be full
UL 1741 SA with FW or FV or VV
- FW or FV or VV will throttle down PV production, takes time, specified in FW or FV or VV over time
or UL 1741 SB
- SB supplemental power controls can throttle PV production must faster
.
Examples: ModBus control for Schneider XW, ModBus, control for Hoymiles HMS-2000, uses WattNode, but needs SOC

Sunny TriPower 33 has SB, 30000 does not.


Sunny Boy 7.7 -41 does not have SB, but it does have some other features PG&E said are required since 20202


I think the Sunny Boys rely on Speedwire/Ethernet to implement zero export (using WattNode?)
Something about Modbus registers vs. registers for other access methods.

Sunny Island is good about sucking down lots of power (low frequency inverter), but I'd like a way to implement limited export at the grid, also limited export at Sunny Island. That would permit me to have GT PV in excess of 6.7kW per SI (pass-through relay limit 56A.)

DER Solar said:

or UL 1741 SB
- SB supplemental power controls can throttle PV production must faster

Click to expand...

For Enphase, the UL1741SB comms are done via network to Envoy then via PLC (slow) to micros so I doubt power control would be any faster than current export limiting power control using similar comms path, which has typical response times of several seconds.

You would think that volts-watts would rapidly curtail output.

The SA modes that don’t require communications respond pretty quickly.

This has some info, looks configurable as part of grid profile.

From what I have seen VW is probably the fastest to achieve full power reduction since voltage is allowed to change fast. FW is limited by how fast frequency is allowed to change on the grid. Both controls have typical change response times in a fraction of second but VW take fewer changes to achieve full power reduction.

Response speed looks configurable as part of grid profile. But per usual hacking grid profile so AC coupling works better is semi legal from a POCO perspective and janky AF from an engineering perspective.

Fast response speed may not always be what you want bc this is a distributed control loop and that may not give a stable result.

solar8484 said:

For Enphase, the UL1741SB comms are done via network to Envoy then via PLC (slow) to micros so I doubt power control would be any faster than current export limiting power control using similar comms path, which has typical response times of several seconds.

Click to expand...

FW, VF and VV does not require PLC at all! Otherwise it would not be allowed by utilities. UL 1741 SA is done passively by the microinverter to monitor grid voltage, amperage, phase and impedance to adjust power production and provide reliable anti-islanding.

Enphase uses PLC carrier communication at carrier frequencies between 100 kHz and 140 kHz, depending on country.
QPSK is 2 bits/s/Hz. Advanced QAM modulation techniques increase the bitrate, like 256 QAM (8 bits/s/Hz). WiFi 5G uses 256 QAM (8 Bits) and WiFi 6 uses 1024 QAM (10 bits/s/Hz) and WiFi 7 will be 4096 QAM at 12 bits/s/Hz. Spread spectrum modulation is another technique highly resilient to interference and jamming. There are many more!

Direct power control of micro-inverters can be done very efficiently and very fast with broadcasting technology via PLC to all attached micro-inverters simultaneously.

Enphase uses custom IC's for PLC and an ASIC for PLC communications in both micro-inverter and IQ system controller, most likely to implement a proprietary technique or one that can be patented. There are many commercial IC's available to design and implement PLC digital communications.

IQ7 Teardown
IQ8 Teardown

Hmm. If PLC is so solid, why did Tigo and Hoymiles go with wireless for their systems? Tigo only goes PLC (over DC, which I feel should be pretty clean) for SunSpec / SunSpec-like simple fire safety RSDs; their higher value systems are wireless. Microinverter PLC is worse because it's in principle in a broadcast domain with all of your neighbors that are tapped off the same transformer secondary. Maybe Enphase's box has an isolation filter or something to notch out the PLC from going out / back in from the shared wiring, like is done for MoCA.

I feel like there needs to be some independent verification of their reliability.

For a bit of context for my saltiness. I wasted a lot of time last year with power line networking (basically PLC). The performance feels 1-2 generations behind where WiFi is, and often has no performance or reliability advantage these days over WiFi if you even know a tiny bit about how to plan your WiFi. So just because there's a wire, doesn't mean it's better than wireless.

zanydroid: Wireless versus PLC: probably fear of violating Enphase patents. PLC patents have long expired and Enphase does not have patents that would prevent publicly available PLC communications with micro-inverters in general. If you read the patents they are very specific.

PLC can be made reliable using modulation techniques that provide low interference and/or add redundancy and error checking/correction. And you can insert a PLC directional powerline filter to block PLC interference with device in the home. Hoymiles uses 1 Ghz wireless in their latest HMS micro-inverters, but all wireless communications are susceptible to interference, jamming and hacking and won't work very well with metal roofs and steel siding. I saw a recommendation last year for panels with WiFi micro-inverters to be installed 1 foot or more above a metal roof!
I know from recent experience: after installing a metal shingle roof, my 900 kHz weather station stopped communication with the base inside the home. If you put the 1 Ghz or 2.4Ghz antenna over or above the metal roof you need very reliable lightning protection.

Why has no one implemented CAN bus style or IC2 micro-inverter interconnect? Need 2 additional twisted wires, a bit more expensive for the cables and connectors if done with a single cable, but would be very safe and fast.

Hedges said:

Can it be told to shut down at SoC, wait until the sun ought to be up, then make AC again so GT PV can produce? SI has some such features, although I haven't tried to implement.

When it is producing AC and GT PV is producing power, is there a load-shed mechanism to automatically disconnect excessive loads that exceed production and cause SoC to continue decreasing instead of increasing?

Load-shed should also prevent reaching too low SoC. I have one stage (all loads disconnected) but plan for 2-stage which will disconnect large loads sooner.



Could feed a charger, of course, but I would like to rectify AC and feed PV (or turbine) input of GT inverter, AC couple that. This allows SI to remain the voltage and frequency source.

Click to expand...

In the Schneider XW-Pro, it has a single "Aux Relay" which can be programmed to do several things, but only one function that you choose. If you add the Schneider AGS "Generator auto start controller" it adds far more control to the system. It even has 4 levels of charge with separate time delays. If you are still at 50% SoC, but for 4 hours, it can start the generator, but if you dropped to just 20% SoC, it will start in 2 minutes. This function can also be used to trigger load shedding, and it frees up the XW-Pro Aux relay to do a different function. Some of their other devices also contain another Aux relay that can be set for another function. They can be triggered by voltage, power, temperature and a few other things as well. But making a multi level load shed and generator start does become a bit of a mess. And I can't think of any way to use just Schneider gear to make it able to shut down from low battery and still dark start when the sun rises unless you have a DC Solar charge controller. And that is what Schneider recommends. Of course, they want you to use one of theirs, but my cheap BougeRV charge controller will work. In fact, since it is NOT on the Schneider Xanbus network, I think it may end up working better in that situation. It does not care at all what the Schneider system is doing, it will just push power directly into the batteries. No data needed, just sunlight. Once the battery voltage recovers, the XW-Pro just turns back on.

I am planning to setup a small power supply that runs off of the 48 volt battery bank to power up the Gateway, Network switch, and my PLC. I will have a low voltage cut on that as well, so if it does drop under 3.2 volts per cell it will shut off before the BMSs in the battery banks do. The PLC could command the system to dark start by watching for solar power coming in from the DC system, and changing the low battery cut voltage in the XW to get the Enphase inverters online sooner. Outputs from the PLC could also be programmed to do all kinds of load shed control as well. Having the PLC adds a ton of options, but it is all on one of us to do all the programming. I have it doing charge and export control so far. How much more do I want to do? I have thought about interrogating the Xanbus. I think it is basically a modified Modbus. If I could make the PLC directly talk on that, I might be able to fake out the Gateway to see my BougeRV charge controller as one of theirs.

DER Solar said:

zanydroid: Wireless versus PLC: probably fear of violating Enphase patents. PLC patents have long expired and Enphase does not have patents that would prevent publicly available PLC communications with micro-inverters in general. If you read the patents they are very specific.

Click to expand...

IMO both approaches require enough diagnostics for you to commission and troubleshoot properly, since they're shared medium and inherently more complex than twisted pair dedicated RS485 or whatever which have some way of plug & pray working 99% of the time.

With RS485/ethernet/etc I know exactly how to sniff the packets and debug if I have to. With the proprietary 1G, 2.4G, and PLC - ????????

My beef with the Ethernet over PLC crap is that it was bottom feeder tier consumer grade stuff with none of the debugging capabilities that you can get even with just a cable modem or consumer router, let alone SMB or better tier networking gear.

Glad to see that there are PLC powerline filters as standard practice.

As for why there is not a second cable... that would probably increase cost (which Enphase probably would like, more expensive equipment and fancy wires = more money). But where it would hurt them is the extra cost on the installer side, and possibly like +30%-50% time. Those cables need to be done up in a weather-resistant way, and appropriate attention paid if it goes in the same J-boxes and conduits as the HVDC or AC. Wireless really simplifies certification of anything that has to touch power. Just look at how many more energy monitors are WiFi based vs Ethernet based, very easy to explain that control/data and power are galvanically isolated when there's no wires at all, and electricians don't have to bust out the code book to understand how to correctly route the Ethernet and AC.

I wouldn't be surprised at all if 1741SB currently only pays lip service on how reliable the local comms need to be, but pays a lot of attention to things like electrical safety.

For an analogy, look at what Tesla did with load sharing on wall connectors. On gen2 it was a serial line. On gen3 it switched to WiFi. Wireless can be done safely and with acceptable performance. Not the BEST performance probably, I bet the charging sessions have to be curtailed when there's interference that blocks the gossip packets describing how much is being utilized.

DER Solar said:

FW, VF and VV does not require PLC at all! Otherwise it would not be allowed by utilities. UL 1741 SA is done passively by the microinverter to monitor grid voltage, amperage, phase and impedance to adjust power production and provide reliable anti-islanding.

Click to expand...

In case it wasn't clear, I was not referring to FW, VW, etc. I was referring to the UL1741SB required network communication based controls.

DER Solar said:

Direct power control of micro-inverters can be done very efficiently and very fast with broadcasting technology via PLC to all attached micro-inverters simultaneously.

Click to expand...

That may be the case under ideal conditions but in reality lots of people have problems with Enphase PLC. Just search for reports from Enphase owners about Envoy not able to communicate with micros via PLC for monitoring sometimes for days.

PLC: Good points! But how many of those running into PLC communication problems have PLC line filters? I remember Enphase installations in homes with X10 and Insteon having lots of problems and Enphase did ignore it for some time. PLC line filters are getting more popular because IQ7's and IQ8's cannot coexist on the same PLC network, the 2 PLC networks have to be physically separated by several feet! May be some of the interference is RF in the 100/110 kHz range. May be their modulation technique isn't all that great because they want to keep it protected under current patents. Other factors are important too, like the distance/noise on the Neutral from the AC combiner to the Neutral/GND bond.
Could be many reasons why it fails to poll micro-inverters for days. Because all updates are done by Enphase over the cloud/internet, the average homeowner never knows if something has been updated and may not even check production reporting for days or weeks or ignore emails about it.

2-Wire Micro-Inverter Communication: I wasn't suggesting running 2 additional wires in addition to the power lines. Something more along the lines of a combination power/signal cable with a plugin tap for power and communications. Would be more expensive than a Q-cable, but save all the extra hardware for PLC or wireless communications and would/could be more open too and completely localized and isolated.

DER Solar said:

2-Wire Micro-Inverter Communication: I wasn't suggesting running 2 additional wires in addition to the power lines. Something more along the lines of a combination power/signal cable with a plugin tap for power and communications. Would be more expensive than a Q-cable, but save all the extra hardware for PLC communications or wireless and would/could be more open too!

Click to expand...

Hmm OK but you still need to transition from Q-cable to other wiring methods, right? Unless you're supposed to use that cable all the way down to the combiner.

Either way, down to the AC combiner or you could split it in the roof junction box/panel/deck into individual class 1 powerlines and class 1 UL approved 600V twisted signal lines that you can run in the same conduit or 2 separate conduits. But you cannot run class 1 and class 2 in the same conduit.

I think there is a tray cable type that has twisted pair and 3 conductors, that might work (and I think the manufacturer has lobbied NEC to expand usage of that specific type to going randomly in walls etc, maybe to cover cases like this). Read about it recently on another forum.

While the Enphase line filter does help improve PLC it's rather expensive at $500 plus installation and some people still report problems even with the filter. PLC appears some inherent limitations.