G100 export limiting

[cojmh]

A tangential issue that is relevant to my plans, which possibly include adding some battery storage - what do you do if it's not possible to physically locate a hybrid inverter near to the grid supply, which seems to be the default assumption afaics, in terms of current sensors/import-export monitoring? Are there any budget end systems capable of supporting some sort of remote sensing - ideally powerline, as running a signalling cable isn't that practical?

I will be putting my second hybrid inverter in an out building that is probably a 30m cable run away from the grid supply, so I have been looking at this too.

I have (or will have shortly) the MyEnergy setup installed which has the ability to send the CT clamp info wirelessly to it's own range of devices. However I cannot see a way of getting this to work with a third pary inverter. I have also looked to see if there are other options for transmitting this info over a network or something else and I have not found anything that would help.

The solution I am going with (unless I find something else) is to extend a CT clamp cable by adding Cat6e cable in the middle to get the distance. I have been told by MyEnergy that this is good for runs up to 100m. The key thing was to use a twisted pair.

Hopefully some of that information is useful?

[patrickl]

I will be putting my second hybrid inverter in an out building that is probably a 30m cable run away from the grid supply, so I have been looking at this too.

8< SNIP

The solution I am going with (unless I find something else) is to extend a CT clamp cable by adding Cat6e cable in the middle to get the distance. I have been told by MyEnergy that this is good for runs up to 100m. The key thing was to use a twisted pair.

I can't see this working too well, but what do I know? The current clamps generate pretty small voltages. Screening, as well as twisted pairs, may help (Cat6e is screened I think). Or some DIY buffering/amplifying perhaps. If lightning is a potential issue, you'd also want some sort of galvanic isolation to protect your inverter from being fried, and one would also want fault detection/protection for the clamp extension being severed assuming it's not built in to the inverter already. Introducing wireless would be an option technically, but that would also add latency. (I have no idea at what point latency would become important, or if the current clamp plays a role in the immediacy of grid disconnect?). Wireless would also make the system more susceptible to interference... Any Heath-Robinson modifications like that are unlikely to fall within G100 tick-box certs?? It all starts to look a bit gnarly.

Based on the generic schematics for SMA battery storage solutions on vendor websites, it looks like SMA rely on a meter that is separate from their inverter to manage exports (possibly modbus based at a guess?), so that might work. But if past experience is anything to go by, it's also likely to be way outside my budget I think some variants of Victron gear may also split into separate boxes in a similar way, but I haven't found concrete info on that yet. So I'm exploring non-grid-tied islanding (useful to know correct term!) with salvaged EV vehicle batteries a bit further atm.

P.

[patrickl]

> The current clamps generate pretty small voltages. Screening, as well as twisted pairs, may help (Cat6e is screened I think). Or some DIY buffering/amplifying perhaps. If lightning is a potential issue, you'd also want some sort of galvanic isolation to protect your inverter from being fried

Experimenting with a small step-up/step-down transformer at each end could kill those birds with one stone. Say a 1:10 ratio, that would still be within SELV ratings then over the cable itself.

[Oldgreybeard]

> The current clamps generate pretty small voltages. Screening, as well as twisted pairs, may help (Cat6e is screened I think). Or some DIY buffering/amplifying perhaps. If lightning is a potential issue, you'd also want some sort of galvanic isolation to protect your inverter from being fried

Experimenting with a small step-up/step-down transformer at each end could kill those birds with one stone. Say a 1:10 ratio, that would still be within SELV ratings then over the cable itself.

CT's are pretty much completely unaffected by cable runs or interference. They are as their name implies, current transformers, so are voltage independent almost. The source impedance is low, as is the load impedance, all that matters is the current. Most seem to work on a 1000:1 current ratio, so the current flowing in the cable will be 1/1000th of the measured current. If the house is drawing, say, 20A, then the current in the cable will be 20mA, and even pretty thin cable will be fine carrying such a low current with no appreciable voltage drop.

[nowty]

Based on the generic schematics for SMA battery storage solutions on vendor websites, it looks like SMA rely on a meter that is separate from their inverter to manage exports (possibly modbus based at a guess?), so that might work. But if past experience is anything to go by, it's also likely to be way outside my budget I think some variants of Victron gear may also split into separate boxes in a similar way, but I haven't found concrete info on that yet. So I'm exploring non-grid-tied islanding (useful to know correct term!) with salvaged EV vehicle batteries a bit further atm.

P.

The meter for an SMA solution is in the Sunny Manager 2.0 device as it has a built in meter which is only good for 60A but you can increase that to 100A if you (like me) add a CT clamp and feed the CT clamp to the Sunny Manager 2.0. The comms from each inverter to the Sunny Manager 2.0 comes via ethernet cables connected to your home network.

https://www.sma.de/en/products/monitori ... manager-20

But yes its expensive, if the newer cheaper Chinese devices existed back then I may well of opted for a different solution.

[patrickl]

They are as their name implies, current transformers, so are voltage independent almost.

Don't most CTs have an integral burden resistor? If so, wouldn't that mean the inverter end of the measurement was of a voltage that was proportional to the current flowing in the CT, iyswim? I'm no expert and am probably missing something fundamental.

P.

[Oldgreybeard]

They are as their name implies, current transformers, so are voltage independent almost.

Don't most CTs have an integral burden resistor? If so, wouldn't that mean the inverter end of the measurement was of a voltage that was proportional to the current flowing in the CT, iyswim? I'm no expert and am probably missing something fundamental.

P.

Not usually now. They tend to have the burden at the load end, to remove that problem. To get around the potential electric shock risk they fit a VDR in the CT, to clip the open circuit voltage to a safe level.

[Oldgreybeard]

An update. I'm now revisiting this, with a view to building and fitting a G100 export limit device, as the DNO have come back with a request that I not exceed 6kW export. Back in 2013 when I made a G59 application (at the time I didn't know how big a PV system we could fit) they originally permitted me to export up to 10kW. When I notified them that I was fitting another PV system, in addition to our battery system (which is set to zero export) they've come back with a request that we not exceed 6kW export, if this is possible. They've said on the phone that their concern relates to several other PV systems having been installed locally since we installed ours.

Have to say the engineering chap I spoke with was very helpful. I asked outright if I could just fit a current sensing relay to the existing PV system AC output and use that to isolate the second PV system AC output if the power exceeded 6kW. He was fine with that solution, said all they needed was assurance that the export would be limited.

The simplest way to do this seems to be to fit another current transformer to the big PV system AC output, and use that to switch a double pole contactor to isolate the second PV system from the grid. Seems this may be easier than I thought, as I've just found these cheap current sensors: https://www.ebay.co.uk/itm/304631567636 They are adjustable from 2A to 30A, and seem ideal, as I can set it to about 26A and it will trigger a contactor directly.



These are self-powered, from the current flowing through the sensed wire, so can safely be tucked into the consumer unit, with the line conductor from both PV inverters running through the hole. I can run a two core cable out to a contactor box, probably with a manual isolator switch as well (just to be able to isolate from indoors as well as outdoors). The way this would work would be that the power to the contactor coil would be switched via this sensor, so if the combined PV systems generate more than 6kW then the contactor would operate and isolate the second PV system to drop the export to no more than 6kW.

Looking at the specs, the switching capability seems fine for a contactor coil (it's rated at 240V AC, up to 500mA), and the switching time is well within the G100 limit, the spec says within 200ms (G100 is within 5s, I believe). I can easily make it "anti-tamper" by just fitting a seal over the current adjustment after it's been set.

It's not perfect, as it won't be measuring export, as such, but that's only because this cheap sensor isn't directional and in our case any export measurement needs to be at the meter kiosk, which is outside and a fair way from where the PV systems are connected. Worst case is that we lose a bit of generation from the second PV system in the event that we are self-consuming a lot at the same time as the main system is generating 6kW.

We so rarely hit the main PV system 6kW inverter limit, and then only ever in the early afternoon when the sun's square on to the roof, that I doubt this is ever going to be a real limitation. By early afternoon generation from the second PV system (which faces more or less east) will be low, anyway, so turning it off once in a while won't have a real impact.

I'll update this thread with how I get on. If it works then it might be worth putting together some proper instructions so others can do something similar if asked to limit export when fitting a second system.

[ALAN/ALAN D]

Have used two of those for over a year. They work well.

[Stig]

The simplest way to do this seems to be to fit another current transformer to the big PV system AC output, and use that to switch a double pole contactor to isolate the second PV system from the grid. Seems this may be easier than I thought, as I've just found these cheap current sensors: https://www.ebay.co.uk/itm/304631567636 They are adjustable from 2A to 30A, and seem ideal, as I can set it to about 26A and it will trigger a contactor directly.

These are self-powered, from the current flowing through the sensed wire, so can safely be tucked into the consumer unit, with the line conductor from both PV inverters running through the hole.

If you're sensing the combined current and using that to disable the 2nd system won't the relay keep clicking in and out? If you're just sensing from the 1st system and using that to switch off the 2nd system then no problem.