East West system two inverters two batteries ?

[bladerunnerpv]

Hi folks
I'm struggling to find a single hybrid inverter for my potential system, so started looking at potential two inverter solutions

should the schematic look something like this, I've set it up as one inverter and one battery each for each roof

how does the system 'know' how best to manage the batteries though, are they two independent systems which might 'fight' with one another ?

thanks for taking a look and for any helpful responses you might be able to pitch in with

 

[bladerunnerpv]

Hmmm, 50 views and no comments back, perhaps I've either asked a really stupid question or a really hard one, anyone feel like enlightening me ;-) Thanks

 

[Barx]

Just to give a reply (although a pretty useless one) I have done plenty of PV work in the past but very little battery so don't know how that setup would react. thinkk you would have to delve deep into manufacturer literature or contact the manufacturer direct to see if/how their system work with 2 unit s in parallel. It may just be a case link them with an ethernet cable for rs485 communications and they become aware of each other.
If it was an AC coupled battery then obviously it would just see the potential export as one value and charge from that, but your preposed setup has the batteries DC coupled hiding behind inverters. Let us know if you find anything out off Givenergy.

 

[binky]

Hmmm, 50 views and no comments back, perhaps I've either asked a really stupid question or a really hard one, anyone feel like enlightening me ;-) Thanks

It's not compulsory to respond, and we do have jobs to do

If you have a read through the posts on here you will find a very similar post asking the same question. In short, it will work fine, it may also work with a smaller inverter, you can undesize inverters by 20%, which can be an option for split arrays, although I'm not a fan of that practice, reasons given in previous posts. Have a good read, and if you need any further explanation / clarification, ask, and I will try to help.

 

[bladerunnerpv]

Just to give a reply (although a pretty useless one) I have done plenty of PV work in the past but very little battery so don't know how that setup would react. thinkk you would have to delve deep into manufacturer literature or contact the manufacturer direct to see if/how their system work with 2 unit s in parallel. It may just be a case link them with an ethernet cable for rs485 communications and they become aware of each other.
If it was an AC coupled battery then obviously it would just see the potential export as one value and charge from that, but your preposed setup has the batteries DC coupled hiding behind inverters. Let us know if you find anything out off Givenergy.

Thanks very much Barx, I haven't found anything in the Givenergy brochures on this point (might just be my eyesight! or maybe terminology I don't understand yet, I'm a complete novice after all!)

I hadn't realised it might be possible to connect the inverter with ethernet/rs485! Good point about contacting manufacturer, will get back to y'all if I learn more.

Hmm, OK for AC coupled the power from inverters first has to go through consumer unit to an AC inverter and then it goes to batteries? Any worries with the consumer unit receiving what could be 8kW?

Thanks again for the (far from) useless reply

 

[bladerunnerpv]

It's not compulsory to respond, and we do have jobs to do

If you have a read through the posts on here you will find a very similar post asking the same question. In short, it will work fine, it may also work with a smaller inverter, you can undesize inverters by 20%, which can be an option for split arrays, although I'm not a fan of that practice, reasons given in previous posts. Have a good read, and if you need any further explanation / clarification, ask, and I will try to help.

Hi Binky, sorry if I gave the impression a reply was compulsory, I didn't think that at all, I was just surprised at the 'traffic' on the post without even a "don't be such a muppet that will never work" sort of reply It's kind of you to offer to help, thanks.

I have been reading through similar threads but didn't understand (sorry old dog new tricks!) I guess I was thinking something like this ...

Two inverters, each receiving power from panels, although levels rising and falling at different times of day according to aspect, so each inverter exports or charges its own battery as it sees fit. (ugh! was that a pun?)

When a load is switched on eg. a 2kW kettle then power is drawn via inverter from panels or from battery depending on PV generation at the time.

When there are two inverters how is it decided where the power comes from - inverter A or B? Do both try to supply 2kW, or do both provide 1kW, or some other ratio, if so, how is it determined and managed? [Barx's inverter to inverter comms link?]

If I understand correctly the interface with the grid is monitored via a CT clamp between the consumer unit and import meter (the electricity meter)

So, both inverters have their own CT clamp and a CT measures current in the cable, the measurement goes to the inverter?

OK, the process could be something like this...

The kettle switch on starts to pull power from the grid initially

Inverter A says to itself, I can give 2kW easy peasy and starts to ... and as they both have a CT clamp so both register the need, so does inverter B supply 2kW too?

OK, so now, IIUC, the currents measured in the CT clamps which measure grid input to house will both drop to zero as power is being provided by inverters, but 4kW is now being sent from inverters to the house?

Worse than this, presuming the measurement is not continuous, it's only done, at a complete guess, at 1Hz, and the timing of measurements of inverters A and B is not synchronised, and there is a lag in the response to the measurements, which might be different for A and B, then something like this might happen vs. time...?

kettle on - grid gives - A measures - - - - A gives - - - A measures - - - - A stops - - - - A measures - - - - A gives
kettle on - grid gives - - - - - B measures - - - - B gives - - - - B measures - - - - B stops - - - - B measures - - - - B gives
------------ggggggggggggggggggggggg22222444444444444444444442222220000000000000000000002222244444

where g = grid supply, 2 is 2kW, 4 is 4kW from A+B, during period of 0 system either draws from grid again or resumes exporting even?

So, the house could be getting an oscillating input from the inverters of 0 to 4kW for a constant 2kW demand?

during periods of 4kW from inverters the CT clamp measurement just knows there's no draw from the grid?

Same situation of a 2420 cycle could occur if supply is from battery A, B instead of PV,
but during 4's could one inverter think it should charge its battery from the other inverter battery output sort of parasitically (sorry not thought through this one properly) and so some sort of see-saw effect occurs between the two batteries?

Apologies if I'm being stupid and wasting everyone's time

I have a job I have to do too (this is much more interesting though!) Thanks for taking the time to educate me. Maybe you should start a PV college and get paid for this!

 

[Mrjmegson]

Sorry Bladerunner, I can't answer the question, as you know, it's one I don't understand, and would like answered myself, I just want to say "I love the way you write, you short write short stories". And I'm not taking the P.

I do know that inverters push against the grid power all the time, with a slightly higher frequency, which I think will mean that power will flow from them before the grid, so they are not sensing the grid power being used. It's my understanding that the CT clip just monitors outgoing power, to tell the inverter that the power it's outputting is going to the grid, so it can either stop it, or divert it into the batteries.

Hope I'm right. Lol

 

[johnb2713]

Sorry Bladerunner, I can't answer the question, as you know, it's one I don't understand, and would like answered myself, I just want to say "I love the way you write, you short write short stories". And I'm not taking the P.

I do know that inverters push against the grid power all the time, with a slightly higher frequency, which I think will mean that power will flow from them before the grid, so they are not sensing the grid power being used. It's my understanding that the CT clip just monitors outgoing power, to tell the inverter that the power it's outputting is going to the grid, so it can either stop it, or divert it into the batteries.

Hope I'm right. Lol

The inverters synchronise to the grid frequency i.e., nominal 50Hz in the UK. They use the CT or other device to monitor import / export to / from the grid. If import is taking place and the inverter has energy available either solar or battery it will increase the voltage of its output which will then feed the load from the inverter rather than the grid (or both if the load exceeds inverter capacity). If export is taking place the inverter will reduce its voltage and / or start to charge the batteries. They are essentially a feedback system; they lag behind what's happening rather than feed forward.

 

[Mrjmegson]

Ah ok, so they lag behind, so not quite what I thought then. Cheers John.

So to go back to the original post, how do two inverters share a load?

 

[johnb2713]

Ah ok, so they lag behind, so not quite what I thought then. Cheers John.

So to go back to the original post, how do two inverters share a load?

They will both adjust their output voltages until the import / export is as near zero as possible. It depends how well tuned the PI loops inside are as to whether they will start to fight each other, in theory, they will both ramp up of down until the import export is zero and then just go on making small adjustments to keep it there. If the inverters are identical, it's likely there will be a way of them communication via a data link to make operation smoother.

 

[Mrjmegson]

I do love you John, the stuff you know is amazing.

Thanks from me, that's exactly what I've wanted to know for a while.

Looking forward to your battery update, hope it's going well.

 

[johnb2713]

I do love you John, the stuff you know is amazing.

Thanks from me, that's exactly what I've wanted to know for a while.

Looking forward to your battery update, hope it's going well.

It's been a steep learning curve for me too LOL

Battery system is coming on well, shed building about to start for housing it.

 

[binky]

It depends how well tuned the PI loops inside are as to whether they will start to fight each other, in theory, they will both ramp up of down until the import export is zero and then just go on making small adjustments to keep it there. If the inverters are identical, it's likely there will be a way of them communication via a data link to make operation smoother.

thinking about some of your previous observations about batteries taking energy from the grid, I'm wondering if in this case 2 normal solar inverters and an AC battery unit might be a better bet? Two systems would suggest doubling energy from the grid taken by the batteries as part of their operation. Ok these are not significant in the scheme of things, but doubling the figures you have seen with your system would suggest these numbers become more significant. Having a separate better inverter also eliminates any possible 'fighting' between them without being reliant on slightly dodgy software programming.

 

[binky]

I do love you John, the stuff you know is amazing.

Thanks from me, that's exactly what I've wanted to know for a while.

Looking forward to your battery update, hope it's going well.

it's fair to say John knows far more about battery operation than I do, I've not had a lot to do with batteries, and don't own one myself as my energy consumption is pretty low, we also fixed our energy tariffs for 3 years just before the jump in prices, and in short the numbers don't really add up currently.

 

[Mrjmegson]

it's fair to say John knows far more about battery operation than I do, I've not had a lot to do with batteries, and don't own one myself as my energy consumption is pretty low, we also fixed our energy tariffs for 3 years just before the jump in prices, and in short the numbers don't really add up currently.

Lucky you on the fixed prices, but it will be a big old jump when your fixed period runs out.

 

[binky]

Lucky you on the fixed prices, but it will be a big old jump when your fixed period runs out.

I'm hoping things will have sorted themselves out by then, like Ukraine has won the war. However, I seriously doubt energy prices will drop to pre-crisis levels. We tree hugger types have been forecasting energy hikes for decades, hence I've sorted most of my energy use years ago. Apart from saving a lot of CO2, I can now gloat and say 'I told you so' . It's what has peeed me off for years wth regards to lack of investment in basic UK infrastruture, governments talking about being 'green' but doing very little in reality has left us all exposed to international events, and even worse we will be paying rather a lot of money to private investors when projects could be state owned, and energy supplied to us at cost. Have aread of this, China and France will be laughing all the way to the bank at our cost. https://www.ft.com/content/9a8ecc04-703a-11e6-9ac1-1055824ca907

 

[Mrjmegson]

I'll have a read of that article later, cheers.

I totally agree, unfortunately there is a lot of money at stake, and a lot of greedy people, so you're never going to have a fair system.

Things like utilities, basic human rights, should be non profit. It's disgusting.

The RAC have been doing studies on the fuel pricing, and say the wholesale price of automotive fuel dropped months ago, but we (only in this country) and still being robbed.

We talk about all these corrupt governments around the world, south America, Russia, China, but our own is by far one of the worst because we hide it and pretend it's the fault of others.

 

[bladerunnerpv]

The inverters synchronise to the grid frequency i.e., nominal 50Hz in the UK. They use the CT or other device to monitor import / export to / from the grid. If import is taking place and the inverter has energy available either solar or battery it will increase the voltage of its output which will then feed the load from the inverter rather than the grid (or both if the load exceeds inverter capacity). If export is taking place the inverter will reduce its voltage and / or start to charge the batteries. They are essentially a feedback system; they lag behind what's happening rather than feed forward.

Hi John, thanks, again, for this I didn't expect the sampling to be this fast, it should help prevent the scenario I was suggesting might happen earlier in the thread.

I appreciate the point about lagging feedback system, it's sort of what I was trying to illustrate, very badly.

Have you come across any info on the lag duration, or time constant? I stayed in a solar equipped holiday property this year and from watching a little colour LCD screen indicating solar gen / export / grid draw it did seem to take a few seconds to replace grid draw with solar power when something was switched on.

 

[Mrjmegson]

You went on holiday and sat watching an LED? Lol. I'm with you brother, rather that than get sand in places.

 

[johnb2713]

Hi John, thanks, again, for this I didn't expect the sampling to be this fast, it should help prevent the scenario I was suggesting might happen earlier in the thread.

I appreciate the point about lagging feedback system, it's sort of what I was trying to illustrate, very badly.

Have you come across any info on the lag duration, or time constant? I stayed in a solar equipped holiday property this year and from watching a little colour LCD screen indicating solar gen / export / grid draw it did seem to take a few seconds to replace grid draw with solar power when something was switched on.

I can only comment on my Growatt inverter but I suspect most of the others are same. The Proportional part of the PI loop seems to be around 90% sod if it sees an error of 1kW it will make a correction of 900w. It the rely on the integral time to ramp the output up/ down to meet zero import / export. The integral part seems to be around 1.5 seconds, quite sluggish but definitely stable under wide ranging conditions.
If you monitor the voltage and the grid current you should see a rapid change if you apply a load followed by an increase in voltage with a drop on import down to 10% of the load followed by a steady drop in import down to zero or thereabouts.